ࡱ> [0bjbj ΐΐ",< < +++8c'+I(./TIVIVIVIVIVIVIK=NlVI0'"(00VI kI7770^lTI70TI77Ah,\B0tl+3f T The addition of input weights and of the threshold makes this neuron a very flexible and powerful one. The MCP neuron has the ability to adapt to a particular situation by changing its weights and/or threshold. Various algorithms exist that cause the neuron to 'adapt'; the most used ones are the Delta rule and the back error propagation. The former is used in feed-forward networks and the latter in feedback networks. 4 Architecture of neural networks 4.1 Feed-forward networks Feed-forward ANNs (figure 4-1) allow signals to travel one way only; from input to output. There is no feedback (loops) i.e. the output of any layer does not affect that same layer. Feed-forward ANNs tend to be straight forward networks that associate inputs with outputs. They are widely used in pattern recognition. This type of organization is also referred to as bottom-up or top-down. 4.2 Feedback networks Feedback networks (figure 4-2) can have signals travelling in both directions by introducing loops in the network. Feedback networks are very powerful and can get extremely complicated. Feedback networks are dynamic; their 'state' is changing continuously until they reach a balanced point. They remain at the balanced point until the input changes and a new equilibrium needs to be found. Feedback architectures are also referred to as interactive or recurrent, although the latter term is often used to denote feedback connections in single-layer organizations.  Figure 4.1 An example of a simple feedforward network  Figure 4.2 An example of a complicated network 4.3 Network layers The commonest type of artificial neural network consists of three groups, or layers, of units: a layer of "input" units is connected to a layer of "hidden" units, which is connected to a layer of "output" units. (Figure 4.1) The activity of the input units represents the raw information that is fed into the network. The activity of each hidden unit is determined by the activities of the input units and the weights on the connections between the input and the hidden units. The behavior of the output units depends on the activity of the hidden units and the weights between the hidden and output units. This simple type of network is interesting because the hidden units are free to construct their own representations of the input. The weights between the input and hidden units determine when each hidden unit is active, and so by modifying these weights, a hidden unit can choose what it represents. We also distinguish single-layer and multi-layer architectures. The single-layer organization, in which all units are connected to one another, to form the most general case and is of more potential computational power than hierarchically structured multi-layer organizations. In multi-layer networks, units are often numbered by layer, instead of following a global numbering. 4.4 Perceptions The most effective thing on the work on neural nets in the 60's went under the heading of 'perceptrons' a term coined by Frank Rosenblatt. The perceptron (figure 4.4) turns out to be an MCP model ( neuron with weighted inputs ) with some additional, fixed, pre--processing. Units labelled A1, A2, Aj , Ap are called association units and their task is to extract specific, localized featured from the input images. Perceptrons mimic the basic idea behind the mammalian visual system. They were mainly used in pattern recognition even though their capabilities extended a lot more.  Figure 4.4 In 1969 Minsky and Papert wrote a book in which they described the limitations of single layer Perceptrons. The impact that the book had was tremendous and caused a lot of neural network researchers to loose their interest. The book was very well written and showed mathematically that single layer perceptrons could not do some basic pattern recognition operations like determining the parity of a shape or determining whether a shape is connected or not. What they did not realised, until the 80's, is that given the appropriate training, multilevel perceptrons can do these operations. 5. The Learning Process The memorization of patterns and the subsequent response of the network can be categorized into two general paradigms: Associative mapping in which the network learns to produce a particular pattern on the set of input units whenever another particular pattern is applied on the set of input units. The associative mapping can generally be broken down into two mechanisms: Auto-association: an input pattern is associated with itself and the states of input and output units coincide. This is used to provide pattern completition, ie to produce a pattern whenever a portion of it or a distorted pattern is presented. In the second case, the network actually stores pairs of patterns building an association between two sets of patterns. Hetero-association: is related to two recall mechanisms: Nearest-neighbor recall, where the output pattern produced corresponds to the input pattern stored, which is closest to the pattern presented, and Interpolative recall, where the output pattern is a similarity dependent interpolation of the patterns stored corresponding to the pattern presented. Yet another paradigm, which is a variant associative mapping is classification, ie when there is a fixed set of categories into which the input patterns are to be classified. Regularity detection in which units learn to respond to particular properties of the input patterns. Whereas in associative mapping the network stores the relationships among patterns, in regularity detection the response of each unit has a particular 'meaning'. This type of learning mechanism is essential for feature discovery and knowledge representation. Every neural network possesses knowledge which is contained in the values of the connections weights. Modifying the knowledge stored in the network as a function of experience implies a learning rule for changing the values of the weights.  Information is stored in the weight matrix W of a neural network. Learning is the determination of the weights. Following the way learning is performed, we can distinguish two major categories of neural networks: Fixed networks in which the weights cannot be changed, ie dW/dt=0. In such networks, the weights are fixed a priori according to the problem to solve. Adaptive networks which are able to change their weights, ie dW/dt not= 0. All learning methods used for adaptive neural networks can be classified into two major categories: Supervised learning which incorporates an external teacher, so that each output unit is told what its desired response to input signals ought to be. During the learning process global information may be required. Paradigms of supervised learning include error-correction learning, reinforcement learning and stochastic learning. An important issue conserving supervised learning is the problem of error convergence, ie the minimization of error between the desired and computed unit values. The aim is to determine a set of weights which minimizes the error. One well-known method, which is common to many learning paradigms, is the least mean square (LMS) convergence. Unsupervised learning uses no external teacher and is based upon only local information. It is also referred to as self-organization, in the sense that it self-organizes data presented to the network and detects their emergent collective properties. Paradigms of unsupervised learning are Hebbian learning and competitive learning. Ano2.2 From Human Neurons to Artificial Neuronesther aspect of learning concerns the distinction or not of a separate phase, during which the network is trained, and a subsequent operation phase. We say that a neural network learns off-line if the learning phase and the operation phase are distinct. A neural network learns on-line if it learns and operates at the same time. Usually, supervised learning is performed off-line, whereas unsupervised learning is performed on-line. 5.1 Transfer Function The behavior of an ANN (Artificial Neural Network) depends on both the weights and the input-output function (transfer function) that is specified for the units. This function typically falls into one of three categories: Linear (or ramp) Threshold Sigmoid For linear units, the output activity is proportional to the total weighted output. For threshold units, the output are set at one of two levels, depending on whether the total input is greater than or less than some threshold value. For sigmoid units, the output varies continuously but not linearly as the input changes. Sigmoid units bear a greater resemblance to real neurones than do linear or threshold units, but all three must be considered rough approximations. To make a neural network that performs some specific task, we must choose how the units are connected to one another (see figure 4.1), and we must set the weights on the connections appropriately. The connections determine whether it is possible for one unit to influence another. The weights specify the strength of the influence. We can teach a three-layer network to perform a particular task by using the following procedure: We present the network with training examples, which consist of a pattern of activities for the input units together with the desired pattern of activities for the output units. We determine how closely the actual output of the network matches the desired output. We change the weight of each connection so that the network produces a better approximation of the desired output. 5.2 An Example to illustrate the above teaching procedure: Assume that we want a network to recognize hand-written digits. We might use an array of, say, 256 sensors, each recording the presence or absence of ink in a small area of a single digit. The network would therefore need 256 input units (one for each sensor), 10 output units (one for each kind of digit) and a number of hidden units. For each kind of digit recorded by the sensors, the network should produce high activity in the appropriate output unit and low activity in the other output units. To train the network, we present an image of a digit and compare the actual activity of the 10 output units with the desired activity. We then calculate the error, which is defined as the square of the difference between the actual and the desired activities. Next we change the weight of each connection so as to reduce the error. We repeat this training process for many different images of each different images of each kind of digit until the network classifies every image correctly. To implement this procedure we need to calculate the error derivative for the weight (EW) in order to change the weight by an amount that is proportional to the rate at which the error changes as the weight is changed. One way to calculate the EW is to perturb a weight slightly and observe how the error changes. But that method is inefficient because it requires a separate perturbation for each of the many weights. Another way to calculate the EW is to use the Back-propagation algorithm which is described below, and has become nowadays one of the most important tools for training neural networks. 5.3 The Back-Propagation Algorithm In order to train a neural network to perform some task, we must adjust the weights of each unit in such a way that the error between the desired output and the actual output is reduced. This process requires that the neural network compute the error derivative of the weights (EW). In other words, it must calculate how the error changes as each weight is increased or decreased slightly. The back propagation algorithm is the most widely used method for determining the EW. The back-propagation algorithm is easiest to understand if all the units in the network are linear. The algorithm computes each EW by first computing the EA, the rate at which the error changes as the activity level of a unit is changed. For output units, the EA is simply the difference between the actual and the desired output. To compute the EA for a hidden unit in the layer just before the output layer, we first identify all the weights between that hidden unit and the output units to which it is connected. We then multiply those weights by the EAs of those output units and add the products. This sum equals the EA for the chosen hidden unit. After calculating all the EAs in the hidden layer just before the output layer, we can compute in like fashion the EAs for other layers, moving from layer to layer in a direction opposite to the way activities propagate through the network. This is what gives back propagation its name. Once the EA has been computed for a unit, it is straight forward to compute the EW for each incoming connection of the unit. The EW is the product of the EA and the activity through the incoming connection. Note that for non-linear units, the back-propagation algorithm includes an extra step. Before back-propagating, the EA must be converted into the EI, the rate at which the error changes as the total input received by a unit is changed. 5.4The back-propagation Algorithm - a mathematical approach (non-linear) Units are connected to one another. Connections correspond to the edges of the underlying directed graph. There is a real number associated with each connection, which is called the weight of the connection. We denote by Wij the weight of the connection from unit ui to unit uj. It is then convenient to represent the pattern of connectivity in the network by a weight matrix W whose elements are the weights Wij. Two types of connection are usually distinguished: excitatory and inhibitory. A positive weight represents an excitatory connection whereas a negative weight represents an inhibitory connection. The pattern of connectivity characterizes the architecture of the network.  A unit in the output layer determines its activity by following a two step procedure. First, it computes the total weighted input xj, using the formula:  where yi is the activity level of the jth unit in the previous layer and Wij is the weight of the connection between the ith and the jth unit. Next, the unit calculates the activity yj using some function of the total weighted input. Typically we use the sigmoid function:  Once the activities of all output units have been determined, the network computes the error E, which is defined by the expression:  Where yj is the activity level of the jth unit in the top layer and dj is the desired output of the jth unit. The back-propagation algorithm consists of four steps: 1. Compute how fast the error changes as the activity of an output unit is changed. This error derivative (EA) is the difference between the actual and the desired activity.  2. Compute how fast the error changes as the total input received by an output unit is changed. This quantity (EI) is the answer from step 1 multiplied by the rate at which the output of a unit changes as its total input is changed.  3. Compute how fast the error changes as a weight on the connection into an output unit is changed. This quantity (EW) is the answer from step 2 multiplied by the activity level of the unit from which the connection emanates.  4. Compute how fast the error changes as the activity of a unit in the previous layer is changed. This crucial step allows back propagation to be applied to multilayer networks. When the activity of a unit in the previous layer changes, it affects the activates of all the output units to which it is connected. So to compute the overall effect on the error, we add together all these separate effects on output units. But each effect is simple to calculate. It is the answer in step 2 multiplied by the weight on the connection to that output unit.  By using steps 2 and 4, we can convert the EAs of one layer of units into EAs for the previous layer. This procedure can be repeated to get the EAs for as many previous layers as desired. Once we know the EA of a unit, we can use steps 2 and 3 to compute the EWs on its incoming connections. 6. Applications of neural networks 6.1 Neural Networks in Practice Given this description of neural networks and how they work, what real world applications are they suited for? Neural networks have broad applicability to real world business problems. In fact, they have already been successfully applied in many industries. Since neural networks are best at identifying patterns or trends in data, they are well suited for prediction or forecasting needs including: Weather forecasting Industrial process control Customer research Data validation Risk management Target marketing But to give you some more specific examples; ANN are also used in the following specific paradigms: recognition of speakers in communications; diagnosis of hepatitis; recovery of telecommunications from faulty software; interpretation of multimeaning Chinese words; undersea mine detection; texture analysis; three-dimensional object recognition; hand-written word recognition; and facial recognition. . 6.2 Neural networks in medicine Artificial Neural Networks (ANN) are currently a 'hot' research area in medicine and it is believed that they will receive extensive application to biomedical systems in the next few years. At the moment, the research is mostly on modeling parts of the human body and recognizing diseases from various scans (e.g. cardiograms, CAT scans, ultrasonic scans, etc.). Neural networks are ideal in recognizing diseases using scans since there is no need to provide a specific algorithm on how to identify the disease. Neural networks learn by example so the details of how to recognize the disease are not needed. What is needed is a set of examples that are representative of all the variations of the disease. The quantity of examples is not as important as the 'quantity'. The examples need to be selected very carefully if the system is to perform reliably and efficiently. 6.2.1 Modeling and Diagnosing the Cardiovascular System Neural Networks are used experimentally to model the human cardiovascular system. Diagnosis can be achieved by building a model of the cardiovascular system of an individual and comparing it with the real time physiological measurements taken from the patient. If this routine is carried out regularly, potential harmful medical conditions can be detected at an early stage and thus make the process of combating the disease much easier. A model of an individual's cardiovascular system must mimic the relationship among physiological variables (i.e., heart rate, systolic and diastolic blood pressures, and breathing rate) at different physical activity levels. If a model is adapted to an individual, then it becomes a model of the physical condition of that individual. The simulator will have to be able to adapt to the features of any individual without the supervision of an expert. This calls for a neural network. Another reason that justifies the use of ANN technology is the ability of ANNs to provide sensor fusion which is the combining of values from several different sensors. Sensor fusion enables the ANNs to learn complex relationships among the individual sensor values, which would otherwise be lost if the values were individually analyzed. In medical modeling and diagnosis, this implies that even though each sensor in a set may be sensitive only to a specific physiological variable, ANNs are capable of detecting complex medical conditions by fusing the data from the individual biomedical sensors. 6.2.2 Electronic noses ANNs are used experimentally to implement electronic noses. Electronic noses have several potential applications in telemedicine. Telemedicine is the practice of medicine over long distances via a communication link. The electronic nose would identify odours in the remote surgical environment. These identified odours would then be electronically transmitted to another site where a door generation system would recreate them. Because the sense of smell can be an important sense to the surgeon, telesmell would enhance telepresent surgery. 6.2.3 Instant Physician An application developed in the mid-1980s called the "instant physician" trained an autoassociative memory neural network to store a large number of medical records, each of which includes information on symptoms, diagnosis, and treatment for a particular case. After training, the net can be presented with input consisting of a set of symptoms; it will then find the full stored pattern that represents the "best" diagnosis and treatment. 6.3 Neural Networks in business Business is a diverted field with several general areas of specialization such as accounting or financial analysis. Almost any neural network application would fit into one business area or financial analysis. There is some potential for using neural networks for business purposes, including resource allocation and scheduling. There is also a strong potential for using neural networks for database mining that is, searching for patterns implicit within the explicitly stored information in databases. Most of the funded work in this area is classified as proprietary. Thus, it is not possible to report on the full extent of the work going on. Most work is applying neural networks, such as the Hopfield-Tank network for optimization and scheduling. 6.3.1 Marketing There is a marketing application which has been integrated with a neural network system. The Airline Marketing Tactician (a trademark abbreviated as AMT) is a computer system made of various intelligent technologies including expert systems. A feed forward neural network is integrated with the AMT and was trained using back-propagation to assist the marketing control of airline seat allocations. The adaptive neural approach was amenable to rule expression. Additionally, the application's environment changed rapidly and constantly, which required a continuously adaptive solution. The system is used to monitor and recommend booking advice for each departure. Such information has a direct impact on the profitability of an airline and can provide a technological advantage for users of the system. [Hutchison & Stephens, 1987] While it is significant that neural networks have been applied to this problem, it is also important to see that this intelligent technology can be integrated with expert systems and other approaches to make a functional system. Neural networks were used to discover the influence of undefined interactions by the various variables. While these interactions were not defined, they were used by the neural system to develop useful conclusions. It is also noteworthy to see that neural networks can influence the bottom line. 6.3.2 Credit Evaluation The HNC Company, founded by Robert Hecht-Nielsen, has developed several neural network applications. One of them is the Credit Scoring system which increases the profitability of the existing model up to 27%. The HNC neural systems were also applied to mortgage screening. A neural network automated mortgage insurance underwriting system was developed by the Nestor Company. This system was trained with 5048 applications of which 2597 were certified. The data related to property and borrower qualifications. In a conservative mode the system agreed on the underwriters on 97% of the cases. In the liberal model the system agreed 84% of the cases. This is system run on an Apollo DN3000 and used 250K memory while processing a case file in approximately 1 sec. 7. Conclusion The computing world has a lot to gain from neural networks. Their ability to learn by example makes them very flexible and powerful. Furthermore there is no need to devise an algorithm in order to perform a specific task; i.e. there is no need to understand the internal mechanisms of that task. They are also very well suited for real time systems because of their fast response and computational times which are due to their parallel architecture. Neural networks also contribute to other areas of research such as neurology and psychology. They are regularly used to model parts of living organisms and to investigate the internal mechanisms of the brain. Perhaps the most exciting aspect of neural networks is the possibility that some day 'conscious' networks might be produced. There is a number of scientists arguing that consciousness is a 'mechanical' property and that 'conscious' neural networks are a realistic possibility. Finally, I would like to state that even though neural networks have a huge potential we will only get the best of them when they are integrated with computing, AI, fuzzy logic and related subjects. References: An introduction to neural computing. Aleksander, I. and Morton, H. 2nd edition Neural Networks at Pacific Northwest National Laboratory HYPERLINK "http://www.emsl.pnl.gov:2080/docs/cie/neural/neural.homepage.html"http://www.emsl.pnl.gov:2080/docs/cie/neural/neural.homepage.html Industrial Applications of Neural Networks (research reports Esprit, I.F.Croall, J.P.Mason) A Novel Approach to Modelling and Diagnosing the Cardiovascular System HYPERLINK "http://www.emsl.pnl.gov:2080/docs/cie/neural/papers2/keller.wcnn95.abs.html"http://www.emsl.pnl.gov:2080/docs/cie/neural/papers2/keller.wcnn95.abs.html Artificial Neural Networks in Medicine HYPERLINK "http://www.emsl.pnl.gov:2080/docs/cie/techbrief/NN.techbrief.html"http://www.emsl.pnl.gov:2080/docs/cie/techbrief/NN.techbrief.ht Neural Networks by Eric Davalo and Patrick Naim Learning internal representations by error propagation by Rumelhart, Hinton and Williams (1986). Klimasauskas, CC. (1989). The 1989 Neuro Computing Bibliography. DARPA Neural Network Study (October, 1987-February, 1989). MIT Lincoln Lab. Neural Networks, Eric Davalo and Patrick Naim Assimov, I (1984, 1950), Robot, Ballatine, New York. Electronic Noses for Telemedicine HYPERLINK "http://www.emsl.pnl.gov:2080/docs/cie/neural/papers2/keller.ccc95.abs.html"http://www.emsl.pnl.gov:2080/docs/cie/neural/papers2/keller.ccc95.abs.html Pattern Recognition of Pathology Images HYPERLINK "http://kopernik-eth.npac.syr.edu:1200/Task4/pattern.html"http://kopernik-eth.npac.syr.edu:1200/Task4/pattern.html       PAGE \* MERGEFORMAT 32  NEURAL NETWORKS Dr. Atif AL-nsour Presented by: Abdullah smadi (2004974044) Mohammed bonyan (2004974094) Mustafa al gadi (2004974041) Ali rababah (2004974065)  R[` տyyfybLE hsh.**h.*h.*5CJ$OJPJQJ\^JaJ$h.*$hphQ6/CJ,OJPJQJ^JaJ,hQ6/CJ,OJPJQJ^JaJ,$hph.*CJ,OJPJQJ^JaJ,hpCJ,OJPJQJ^JaJ,$hgQ5CJ4OJPJQJ\^JaJ4*hph.*5CJ4OJPJQJ\^JaJ4 hE|hphsCJOJPJQJ^JaJhsjhE|UmHnHtHu ( C x P d^gd.* dgd.*gdsddd[$\$gdQ6/ddd[$\$gd.*ddd@&[$\$gd.*gdE|   % & , - * + @ A G H [ \ 7 8 M T u v | }  56Mĭēē3hsh`c5B*CJOJPJQJ\^JaJph-h`c5B*CJOJPJQJ\^JaJph3hsh.*5B*CJOJPJQJ\^JaJphh+jh+U*hsh.*5CJOJPJQJ\^JaJ:MNTUXYls"#)*12QRXYYZdt~depqvwéé3hsh`c5B*CJOJPJQJ\^JaJph3hsh.*5B*CJOJPJQJ\^JaJphh+*hsh.*5CJOJPJQJ\^JaJjh+UB%Ts}GHU z dgd.* d^gd.*hiz{DEKL)*EFLMnoð|3h`ch`c5B*CJOJPJQJ\^JaJph333h`ch85B*CJOJPJQJ\^JaJph33$h.*5CJOJPJQJ\^JaJ*hsh.*5CJOJPJQJ\^JaJ3hsh.*5B*CJOJPJQJ\^JaJphh+jh+U,owEJRS[\  }~   lmwxz{=íéíéíéíéíéíéíéíhR|Sh.*CJaJh.*$hs5CJOJPJQJ\^JaJh+*hsh.*5CJOJPJQJ\^JaJjh+U3hsh.*5B*CJOJPJQJ\^JaJph3hsh`c5B*CJOJPJQJ\^JaJph3z{}p K!@$%%&%'{''()))++,:.gdR|S & Fddd[$\$gd.*gdxkgd.*gd.*gdQ6/gd.*gd.* dgds=C%-OWH O P!Y!~""% %%%n&u&''''z){))))+,,,,1111111ӭxtpthxkhR|Sh8h9N CJaJh8hgQCJaJh8hR|SCJaJh9N CJOJPJQJ^JaJh9N hR|SCJaJhR|ShgQCJaJhR|ShxkCJaJhxkCJaJh.*hR|ShqCJaJhR|ShQ6/CJaJhR|Sh.*CJaJhQ6/CJaJ,:./11112,5.51535K5N5P5]5^5Wkdh$$If0$BB t634BaytR|S $$Ifa$gdxkgdR|SgdR|Sgd.*gdR|S12222;3?3333444,51525N5O5]5^5a5c5d5s5z555556666677ɸɧɠɜt_(jnihfhfCJUaJmHnHuhR|SCJaJmHnHuhR|ShgQCJaJh9N CJOJPJQJ^JaJh9N hfhR|S j>hIhfUmHnHu jhIhfUmHnHuhR|ShR|ShlKCJaJhlKCJaJhxkCJaJhR|ShR|S0JCJaJhR|ShR|SCJaJ#^5a5b5d5566667779999:0;==== $$Ifa$$a$gdOMgdOMgdOMgd fgdR|S$a$gdR|SgdlKgdR|Sgd.*gdR|S7.7C7S7T7U7d7e7f7g7799=== > >)>*>4>5>\>]>@@@@@@@@@AAUArABBBDDDDDӾӳӳӟwh&)CJaJ(jhfhfCJUaJmHnHuh&)h&)CJaJhOMhOM0JCJaJ hfhOMhfhOMCJaJ(jJhfhfCJUaJmHnHuhOMhOMCJaJ hOMhOMhR|Sh9N hOMh fCJaJh fhOM,================>>> > >>>>>>> >Ff-FfΩ $$Ifa$ >#>&>)>*>+>,>->.>/>0>1>2>3>4>5>;><>?>B>G>L>Q>T>Y>\>]>FfJFfFf $$Ifa$]>@@@@@@@@@@@@@@@@@@@@@@@@@@@FfFf $$Ifa$gdOM@@@@@@@@@@@@@@@@@@@@@@@AAA AAFfFfg $$Ifa$AAAAABBB(BDDDDDWEYE[EEEEE$d$Ifa$gd&)$ddd[$\$a$gd&)$a$gd&)gd&)gd&)gdR|SgdOMFf% $$Ifa$DDDDREXEYEZEEEEEEEFF;FF?F@FAFBFCFDFEFFFGFMFNFQFTFWFZF]F`FcFfFgFFf"FfFf^$d$Ifa$gd&)gFsFyFzF}FFFFFFFFFFFFFFFFFFFFFFf(Ff{%$d$Ifa$gd&)$ddd[$\$a$gd&)FFFFFFFFFFFFFFFFFFFFFFFFFFFFf/Ff9,$d$Ifa$gd&)FFFFGGGGGGG"G%G(G+G.G1G4G5G;G]J]]]]ǷǷvcPcPc$h fhsCJOJPJQJ^JaJ$h fh fCJOJPJQJ^JaJ*h fh f5CJOJPJQJ\^JaJh fh f6CJ]aJ(jhfhfCJUaJmHnHuhsCJaJh&)CJOJPJQJ^JaJh&)CJaJh fh&)CJaJh fCJaJh fhsCJaJh{CJaJh fh fCJaJVkXlXnX~XZZZ] ]"]:]]^$`e`aPbRbcd dd[$\$gd fddd[$\$gd fddd@&[$\$gd fgd f$a$gd fgd fgd&)gds]]]]j^u^^^^^%`*`+`=`e`h`ͷ|_I|,I|8jFhfhfCJOJPJQJU^JaJmHnHu*h fhs6CJOJPJQJ]^JaJ8jBhfhfCJOJPJQJU^JaJmHnHu)hajCJOJPJQJ^JaJmHnHu$h fhsCJOJPJQJ^JaJ$h fh fCJOJPJQJ^JaJ*h fh f5CJOJPJQJ\^JaJ*h fhX5CJOJPJQJ\^JaJ8jC@hfhfCJOJPJQJU^JaJmHnHuh`i`y```aaaaOaSaRbSb]bgbbͺxͺ[E/*h fh f5CJOJPJQJ\^JaJ*h fhX5CJOJPJQJ\^JaJ8j7PhfhfCJOJPJQJU^JaJmHnHu8jLhfhfCJOJPJQJU^JaJmHnHu)hajCJOJPJQJ^JaJmHnHuhajCJOJPJQJ^JaJ$h fh fCJOJPJQJ^JaJ*h fhs6CJOJPJQJ]^JaJ8jwIhfhfCJOJPJQJU^JaJmHnHubbccddeeeeef f!f)f2flffڽڪwaڪDwa9h fh fCJaJ8jޜhfhfCJOJPJQJU^JaJmHnHu*h fh f5CJOJPJQJ\^JaJ*h fh5CJOJPJQJ\^JaJ8j^hfhfCJOJPJQJU^JaJmHnHu$h f5CJOJPJQJ\^JaJ8jRhfhfCJOJPJQJU^JaJmHnHu$h fh fCJOJPJQJ^JaJ$h fhXCJOJPJQJ^JaJdddeef flfffpillmmmmnnoppAqgd fgd f dd[$\$gd fddd[$\$gd f$ddd[$\$a$gd fffff.h8hyhhhh9iCioipiqiiiijjjjjj*k2krl~llllllmmmmʿʿpeh fh`cCJaJ(jbhfhfCJUaJmHnHuh fh fhsCJaJ.jh+hf5CJU\aJmHnHuh fCJaJh fhCJaJh fh fCJaJh fh f0JCJaJ.j^h+hf5CJU\aJmHnHuh f5CJ\aJmHnHu$mmmmmmmmnn"n2nnnpprrs$s3vDda   A?ØreportPicture 14http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.neuron.jpgRM=v9ަQvU )=DF!=v9ަQvU JFIFVT-COMPRESS v2.1 C    " ( #$&'&*-*%-"%&%  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzVTC-2.1 YUV ?>/l¦'jX)hF8T a{zԁp{~Uv>Awqk!9FO?hRYn\$]ZQ^M='=S>+Mc^=e?[޽h 8|OhCʆEPY (h0[\M,VI$$, M\/>,xn(ͥxeV fc i/pqSvpH;cץMsןm~>Cڕ6;#hwimXX1ZXA^{qpWqJ,|#ዝJhƇʌW'$5+6n^Ov]qIWo2{1$;՞gD@MwS%]7+)p6H.C.X~SKO ma֥gjWG#ҾRt"- A˗[sc> f?4=/LǝoE#;ǞW{`צi^occoo(UVW|[c3$|kd{E5sz,50[dDOl֏qGqGqGqGRGqGqGqGqG { u{xkOK95]EN~ݩneϨ-@ tzQvJ8JeXrt{WyOZ}ޡ=ݯo'v7 <#=m|:65쨲c7`Ak>Ӵ֬dfƳā=o3ߎ|9|2im<J+MA P1ZA_ʺ?U֭mgI:2`ˬZۍ2GL[kW* :x.Nsܧe^ʼ|9'%^]*7@ m>W]6=4J6u-r}g@u}Q|[o1OҠ?;DG6w${rO+o#UN=WG#s?Rz^ u-}V5j˝_Qm 5Ӣev. 0n_p*v!U+4JrE%X~b&R6u˄O?Bq޲osopYj:gd0>Tڦ'FnAjUMk3KNN{}f=!.pzWrS֧-tvuk,sD*ָPk2bo*دB c`QFQֵ3T6v}Z.G64ɝi/t )t-b]Ҡ@Uޠ9^ˤEmUF 0:V@IGjZ(1mcBKYhV)y37a}@-U=KXtkuW2jOIf HnWBa +:!2KKE ((@;dRE㫧]W6 E"]b"E!=c$\/? :n!f#Ƹ}2— >?|pnNԭm$Nʻ#߷z߸5G *p ޵2ڠF r?* o? Y9$Sȿʳ<-x?NL<9p.'{d'+Q[6 T5kt-bX7*~Obxo/·:?`gxcx+|kb<':PX Gue[f[mBO0-Ʋ5_ i&krdr}wxOůo}{+gϸ?޷&@]ڔ:/+X4ٟe& JU2Aޖ(."{8bR@$ׇi?$]IG̚j yqp sOQҊ4o i}B qRX9'S|1k諩)r-] DdnHsYӴj^,M$VwR^(X ^O\b?eiz#VW.ys35ij-S"X䁵s1[c^R>0|F 1ψ2,14u(.:=xpxYmv/JHF$(w_`U!w属)9$EQEQIGA((}&k^T;R[Sҷ*8UO[Ҵ;eYl#<c"̚v,^(թxg6* ?+ZngO\!6ke? x={ ?mbG~1ƐıāFT`_Qǥ? ^+ЦDcr)Gyu[7_o_:V9>YS,#}ԭ!I%h\e^6 =AS#i4,*'y5[,<1iR>OVN!Dnǡq]/|"bE80PN@y0:óh ρ3NtA^g/4vs:OUq&2Wh6;Z "-Pk?o뇐soFx!^UBxJdCq EqIuGPA)FҠ=8TWem}J4j۟]iw-FcTW>ލkK¶"ј {zϹ4+M2D7cd]wzQǥzQǥgҎ=(Ҏ1Ҋ8J8J8JMݵn ip,Y $֜kn\HE,w&OuiZ\2D9u˲~Χ#wš7t[xiO#7<¤qQ/fu5ZQ@ `WWi:Ml3+@k>XKW; U$[#@+(h +]B⸁9T2onԾjJ&zKZI;%|=X/6uOr֣>)4Ӆ|4(QH89Q_&ⲇ:iJ0ijOmYliLvo-HZ/i^&ғPЯbbW|} ץn+~2\\xvV\+l"1˟ʻ-HG,#D~w:Lf$7n88futXI,wp }jvs3]h1s Y[b0qoAx?fk>,t#Q tC2赋0FPG<$HZEASVM4ewla1{k-/*}ej?[xZ}7WɑWc.sΡOK- [6:ٜ~5 K_Ze֚ͮ?:+ğYڇk ɞ OӵDX>.Tt>Re3@?T>hɟG֣nV6*aW'tx_ڷ'9;~.%*+x JO֕m0jOFxXԐ~Frߨ^*  >qԃh@lj۬EL~6X&%X_Y}Cm0^ ?i@c?4K_xEn3VZ oǞhe?/O%SQ)V4JcP9LV¸#Ҫk:j723jhVrL`}#_gQԬ:[N+khV >P_x+K]ҒU!~izz>>abYeq3UC%G*@צ=h$h F *q\)888'˩\s0M9fs]=ǵcڌ{QjBQMG G-<-|g<,6C~z->7}vh}(ʌ}s[|hAm#=3~d\> kEze$AxIo>[6?'һ08QjhtP <Ӏcڌ{QjN>I퉺5 [GO \= `i4?p.-t6Tߦx 8i3vJ,aIj1F=ǵcڌ{Rc.=ǵ||xҧSm̩P>f9|6&em};NqkqhDZƌ{1hDZƌ{e,\Ž^iT3d$? OJ:u{%ܑ9@]=4c'㟇/ k4jae!l 2uiƁ#P8Fǵ.=ƌ{1hDZƌ{1hDZƌ{1hDZ3ּ}O㕑4:eLy]H$@F8'z=vPG(H8DZi +kh'Yf$h!ff<'k`QF^P-B+hw+\3j+'R5?WtوǙqiMqT隥@E;\(zN`QF`V5 ~Vem9>zW "_ϏCh-/i2ֶA>Oֽ 0(x*:| q,Y۶-A:dU*?4OƱ<+Oܕ} *#Ѧ5X&h~Vh0m>5 xTtQ"d}_>.n4ǹU##BNОO֓þ~"3 Rè貵m[@n>YG WHWC/ӾՠEr8#yp( 0*+llfp†Gc*I+#~$xNZk>ц!X13 0( 0+ӧˢG%u`VGn[[[A,J4Y3I[QEax ^ho!PcT+/nτij- =x"'-de3(:(y⶷ydHA Ԛ"<'܂SRwW9TU TԴQEvetԴ!`GMwV~i%uȭ-bg>w>°\CR" 2mf%I=WX`]F_@>W/Zv&]OaM e/Gj|3%Wl~$\rTjM;_g-o[6ٰG]9W['Ӵ{nY19]Ɲaov6QXAUFQEQEruh+W/*!}Ed-d=s+Wiz]Ca[Gmk(EQY!mMGYX!N)Oa\D:$²sO= R{pDŽv^"Û=Y3=QIߞX{V vBn'@jލ 𶷮 Gmˌ-ީ|Pe;4k.u3c=>MҼO=ƉsC.0ĀsjZ\.e"jEAuek{Bʁ# N8QHHQYKV{4\ذ?ZJ((nb8RpkÞ6|Gbv%1:6C98zUC@Ӯ :hT[q,0~`zovO]⵸ϓ3F\usjвԬӮae%ãb85⛻}?A7w[[A<4"E 31$ǥ2i8fcw<)]E$9]cS[+|Jɯ'5-|u# Ԛm~B-1,H0Y$dMj:~M6@"T}?^ҦӵH']E*Z-iqOwEqesm<9x߮O רC+ ;^=(Ҏ=+.taм/n/|E~mGY=g7M_^}U*:$1WqGqGqGq\.E:t7mkv+T5p;PUYJ<^m؟_?`@<=Mj1ktYaWGqA 2j kRuhW'' dk׈xzUҟ|3Ttmv`x"8J8J>'w6z5mr_?v|}MF,%6,K}GqGqGq\玴GMKc w,wSCՈ81޹figZrhH!|AT_u YkUs'k"c~[8I#*<A{]cS4-l1 P' I#]4beDaFFaSxA7mmo<4#E 31$_Qum2{B/6t)"dL,m2 .-a] h~5xKݴwkk ,7@OnM]%h卆CFE=UK1I=GB \-*//O%?zei=-xQ2^@9Ͻt4Sdu&y*('Px%Mՠ:~eiDǐ[=SBվHǂkye?zHIc_$o<[J-[OrM%X9\Ɠ<~rFPjNj|kl䲪s[>}tQyc);.m$?uGnEq;KQե[ esH "öZUHR>F3(((Mݽnx%i\"D2fcI5GqGq\čjKVzK2j:X[29#J6%אC[7J@ZYhz[H.n/" IW8-OxR+Ie}=`Tbߌ|R[k K[?-Xm[-..yAz?qGWSԭ4}.{Fd # k'baCZHx2Rx2?|B)n \,;dAM8oD?,ٺ^#t :I卍K ""/@xFJ_u_q!C,cj<)>nq·l֓Ր1\i#SZzN:Y QzRx~^^3'ռ9mɢxǯ+B!ע𿉡Ɇ< IϷJ8/~!sM{==*3W}/qGqGqGqGqGqG⛻k O,J$Y3I|{Qj1F=ǵZq|t4JQ34z1F1ڸ|BM'Z?_C>Y8<TS;wD}5F=4c*xxǷ*KiZS=8IG^cP\Z^& XS9d0/yf{)r3$¸ZWtSkTFv66vk F0+׮a=7Ye[̒Bpe;$]`TwouK U`z+?ko{E]Iqo=2*]CහZnw"Թ{ y횹]Ou7TR? Ow̟^㱮 0+Kt۪lR?y!UɮE-- LPZı `1] 0(wškU2 'nR/߇o}7R_@z)$2f`ݍ 0( 0+'v~n流 Yp,Y $z((7|v $rO ,+OZy9-SizDDq/(8`*G߇n5P&KIaJ9#1𯋮_V >McZFO:뫚Ҽ/ug#ZR"t(,nIEQ\mh MeA!V_|?xHMlr#sPg@o ?S!F&pvMv'&mi=^"ƾbꡱ= 6VR8Pz*k5lj5ˏH~ :T.1[{lWwEOmo]yR-`yu9^wCS( Yp,Y $z((\iEcr-nc-J-|;u>uu>\>tWIEq5wok{:doۭv-c-QUm4+ [gx)N*QEQE_ ?o$3j,;Ǧk+[hඍc5 tTQY0xGöAҢV'KHz9gEWQҬRRqv=4CcmA_)pduTSMXj)XjU=&kv6 &(02ACϓ W#R A_O#fDu_T`:`[```QEQY)w}q,J$Y3IZ`QF`QF`QPEm5,EQwl 0( 0( 0( 0( 0(a-Ci+^O>8 O3Z`QF`QF`QF`QF⛻m?A7wmmK,8Dd,ǀzWA*Dd   A?ÚreportPicture 15http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.synapse.jpgR(B+4A2(>F(B+4A2JFIFVT-COMPRESS v2.1 C    " ( #$&'&*-*%-"%&%  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzVTC-2.1 YUV ?_o|HKrn4oilܡ8bJ> ji^ M= ,^Y8?Jޓm;K2 B_ (s\GG5kȰɂ2WA'EQEQEQEQEQEQEQEQEQEfM= r,?6A Aod}y‹5 /xj6Y6\NOP ,ׅ|Uusx!.⑓]'􌞣uJţYl'HrʛEuQEgk>"<= >q}G,~_Z-^(s%g BۺW9/ep‰h"Rb1F,rGq^E xvKćN [ Ҵm㸷C,mq7Ƶ#%B`@؜ j*J}MfajZc¤xZSJ-Kc sJš ͵a 7\N/׎jmB>9FCOF<zSxn,,H+3HfBpJ4>t=:-QI*٥d Wr:՝z~e^^A,s<8ʩydχ4mJ,`aݘ5EQEQEY,u'U+N CLׄ6O+zrT0lz[j|[kđ4٧ǠX[ᦊcMwF4Uڒ[f2 ?Nk((PӮ|1e?ᬤE 'Ԏ#?ku w51ְRq7RZXG7?5yO wTu?گ¬6_UOQӑ&),Q⧏CIi`,ۆ5j/+ 7q{=~@ԫB|:2ڔ{Iq{pF "_,4|>vھ&QZ\ hO0\~x=>%)#:8H?c =yZxۀE>?JԿhZ)~И?Yp<$8?ʹ4gG"1:;Xj}*?B퇅a2^_!##y|W]q$ X#0tI-S |&uD[8O+zH32i&s}=I8h((4{ h2O񮍧^' HYYHA{`\(Eյm&(##rOZ|?3kP[jߌt Aa1ҵ9䰁u.&>`VQHQOU*7~y%4,mѠS* ī<9pQ<.< hDOo\X?"{šZ劫'R9^EMQ?.aL?>7v#췓pKZl#dZ >[j|˄:gnWj炝p'-,LյE/ᢸ\J*?e= SZSat|qM@xi)ۿ)'խ ߲6*z/Mo>oYro:0},~-x,{b ݝWv|OcR_%eԼy؟28PҊ(f@3\÷{ IrK_EtQEQEQEQEQEQEbxVѼ'=Ɠr޻ !wꀟo#_33k~W 7=t]$dXE.G_x7W=?Ц /""O4r\3$c!Ga$hzڎY=REg,A = ?_&X )VN<;aEbxPIujEJqFH弔q#𮢸mo5-=%^=O](t*2C_>HoD5x_ֲ 'S;f':gZisYZ"ck:nRA譡߬:->>oH6^aYGZtQE|]cAxtm>"8`-bG\Gb=+(k(!B@_iKot1 "¢;eTƺZ((((##^mo*URTz85QEQEW#/o<7≵ D0:\X5xKqXXH9b#!'vS,5 5xswSYO?NXo\GacqEv:Fgil"xjO=ItW%4"Q{8>zm9.vگ K5ͺ]my ,G#kg^.lnn>m,r f ~ r@dj𞿦! ͢Zs!]1׃^gF{m D*[V|}y5KFnmYl`#Ӑj/ pZ ^Ykf/ǐAPs^]cAxtm>"8`-bG\Gb=+(/0iw2L&g ,KuݎKIY!i8"Qi2ȨK2 vOXj6q\i2 a-3'6Swf'G(o{)5VYDѮkBԼ}OOeY*ӷy&@H)OX,-zG Q ((+ƞA)i'+VnUszj~1WT5NA)^?둚h( Oxį jڥGD-VXJ>[zS@nG⫻+y&K tbߊF<}w%B/ o5=c!uꚭ}⛙WY6Iclqeyr'6>MmǧSj-m-mN8*KEQEQE|]cAxtm>"8`-bG\Gb=+(\BBPb1\yKkBŤy,ܟXدt΋$l+ {|4ffKslyI]QEQEQEQX^-twݼ[mOQ޳mgIXGͮ#I]ױhXuks/Mex`pRXO!EQEQuxtw*FBՈt](((+׼˨IQo.e>pٳQhƫe-Dev/ @"z(> X}l2DxR6ɞą>-KKϓqʙ2*z((((PӮ|1e߅%.~ҰBJܢ((* g}i{n8W/XoNMp*Ѿ jαm/J9+Z}y:}qX vvu jѮV;{Bwq0U`p7dgx1"6c@<1syoy i.&;=ly ;&㩭 jK֭J#RYHr:9뢢+ G ((((i!FLUPˠx^i#Уr{שּׁlX`Gx_MVr|)bV\Cƞ *xM|j={Hq&|?@ivmɝdWAY i+$]LݕrMfkIU;˕ڒտڣV}PxA繵O.>pmzV֍X\w\,N:GB ;V4 LX1FS+|KxOhX7O`AEíK\:뻩1i=C1&­zKgT=VWlah^ ztsY>H#کx~S_tx'*I8Y#o I7X4l~DӬ*xBWK)pI;a1xVF6!FAI wKIvX|7eLpZ_P=N69פVݦs[ #л|~ե𮶸r3Y4Z>-~FUPՌүj8F@Eg3&+^t> ^<Ӯ|1e!qn~k3p3We,1OI{[ttK[{r.m]o֢6Zv#ݷ2$o\4j2f?h" 43-)4Gѣl2"hPӮ|1e,F=?yk(ꇟǥnEQT^A5Rt.5xJ:Ŏ-s%6WSSlcyyj&ĺLHë9U' (cZ ӡ)jM.S\qEQEH|?V&6=}*/|H5tmOHմMRef r.ѓ6+|%cj14d2?"i_-D]kSDW:D=ڢPӮ|1e]틟'"*(+ : SkڻZFʜ{UmOM4J I5=9|:-槥jw&"PTn+h(((PӮ|1eV#i֧V/ s-˶Լc8ueAYY7ڠPH-F> #CPYPe,ԢBoFTBf p8V߈znEӒ6Ֆ?dm$yqW7j`vye{mEwŴ9c`SQEWoZtaϻ?WeEQEQEQEQEQEQEQUuS?~?&~ᮆ}3kcQc i=>3T xtWAXwčiՍۭGUbÆS5+ sp1ۂqUv>Fkhe28fcUkMI#躪mpnmKˆ]=r=ol]iiSg8I-PFk{(aa# -`{~BC(0śzz|3͡rXe=ŚkGZtQEhO_aghOk(((((QAq\L5eܱ:#$N_Q$CO%r#j0ԅQN^MF|޸S\_>)J=֣:M Q7  lzeiA0@o/@?⧂o"8 18q]Ur Ӄ0qXVlanB8ђѶ& j`xW{O3IQǽ5? zqxȸ [h*F!864&wKhfB2}W9vH5[`1'G#57W` oݏWaEW<5.x{Yӥ"7A 5 -~5+ESnBpA=ɮ(((((((+y)륨 3He2*Z(u~Y,3}c|;JvQkDGJH6bl=>}Ƅ\|Cχl줘)d >tXsYqb[״O||ǑڵNGqq<\RCSߍHk'BA[^It#_]ing̶.߶-խ4Rm'_6ޅaa“G8S`cM_(((((+ ދ6D2Go Y!(njї0SƏPidǟI1;%- J}cCf6iF[s}+9Fj:]֑si} qGZcuLJ &MǍ?n|Hp``ei|w15hoڮL`aAhSݘ> oV@)T#M$hdoҴ hZLnE6rfHOs?jE|:iÂ]V|>?utQTL۪icBG7&γ}|\[7G9 /n-a]Jԗy{PeaA"(((((Eм&H]$=19 gxY^`IԃW&;;Vtϩjg'A&{XW'?A[rFDʊXdBݼu֚tX,݉#9¼YQEQEЧuyLV]#'֓Am9o4 BLM:8((z=Os!BHcu 9G*|C9\EF=/n|].,S4pn;\rGji?Hy}^AwOۦ*7$Z (3zKq趖 k-ƭn:s; ӸlmOӦ xtM~(;ܟ[49ӡaWFB=W:ǧAǑu bYĊ Kjֳp!dsc=w5[Q{Pe6|yܿγĪmkLakW7+}~OUC 卷iA"%c=+Km|JMOK(ܫ |?#Z1  ͉f|ec]m=Cn8bJ(gQ-Z Uܹ,Ozĵdž77+X #DžvYLqE[W3stE`cl'mռ(?o.eWqw>Xmco0;Itbz((+ՆymMYjܼD= 3ȡK5 j ox_?75p=,դkhH]8Yƹ}wȞl RyڪrF'wQ\;}ZKյ+3!WڠI rj'3CSJ|s PI- qa}B#h`vFBpǁ^/o1geOBk sI;l|R8~tu=7z4Rc̉Up Es>{ⴶ h(cP *:S袊+ƾal5ŝ298+ᮁ^^Z5 "(((((+U-]MQE TA\ dށ{'-qL`-Aj m/& 4]~ Z|[$OO1M21қ n|+x?d ڒOG1uU:\Ak+!Dj:O#XO쫆pζ_D^$t\XiZ3\<7me't,̗Wr\8"(x uQeajtM4l^A>jQEU}H]d gx. "ZF]Jv=+f(((((+1ݜ.V ((:edd>bWQEQEQEQE,[dpѰsqZl`@'hռ"G4L}VҒixGOҥ(((<;Ir+8xYHH=iÈrծIy{ ^ bWSE?m<1+Ðiڦ]͆Ipd1=h{ TMum먢(^"qɩ;'36ɧ3k{$($ ;ZN(((lϵh'jOi&m%ލQղ oxVI:zufh"%٧ۃpS׽giZy8 _ivz]:Z)L -Ily]^}XOU9 b 1O y((}6mmsq'0`J:Ú!k= ^+?GYOo~ʫopNq\Wy]Co m킭ɎRB'ރ}aswqc“iµ  ː8^=k?V"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz ?h ( ( ( ( ( (.nPӳݱ99 ǯS]d5,.?!>` G͎q@-\26Ҳ9$ϯ$w9ˆ1O9R;PK<1=~sF [=r8=88g~q;nݻou9 J48Գg [ ګ[Zᶀ z}2$iHsfܽqWX˫Yyc \Uy݃O_c V p^hiY@ր' ( ( ( ( ( ( (^]V?z Y~]sp]udZG ky]94@P@P@P@b!W>dlP99$,jR58-R? #P@P@PWܠI窟Pz9 g_pz8hɨb616S7oN23@ b\3+|.N=.ynSȈ5t`H[gHP-eThF{ hEdEt`*r( ( ( ( ( ( ( ([32,F=tҴA+ :bE ikQ@%P@P@P@5dFGP 24Srsc.L}꽀@bhZ7r6oH̊h8X`z (o}0tzp'F9?޸>g= ʀ.]C 39PG,AF9i!&7o\n/mv뜅?L;9oqe*}<ǟQ@@P@P@P@P@P@P@PYfRqϱ\RWb(.>qAK/"\RWb(.>qAK/"\RWb(.>y1jԭ.]@%_Ӟ9=(ڤFF3( T<<g5 x7gV?6Bcl0zҀ#Mbɤ^D LB {Pjy(%_cA8`8@P@P@DT0X3r=x϶;PP@P@R?y v g:( ( ( V_dW@v ( ( ( ( (BEgMsP|ͭqԃ8{s$LU(O;s hgWFx]E$n8yǠ ..U$X(H <04 IKlD##?mX7T*ZWR(:w&v;C?HPr9LDa-y%  7x<Ӯ!Z99A =  8@4h)7ɭE6eoo΀.@P@aZ @2@@ iwo{d3ON3yhz(>/^.:$jv\FNHGu2[ DSڴF~Ls@+HHG-\>Sۜ ogpڍ$%!F}Y8CW ?ͻ#N40ؠȁ]#BOėWh/baXl u9Ghφ7ȣk#hP@P@P@P@PymǕeҀ' ( (&h´rGk#€.@QM*!l,m r`?4Bq$œЌ眎(  Cqt4(3 `1HZ09RU G$ӵV]8xb6$l;v Oz|i@g][A{BC YdPK2<I:P:P:P:P:P:PZyJA88?Lҭ4LVݍNYI#@h ac/eB=NqHs@ 5?a%(^+;vO,*H;n;,y'~1@ -Y& nہ~s|z~}ͰT7#2wt$r;Ck*%%Sq8c;vv[+JY  ( ( ( (8Vx9PP@P@R?w w_ W?zP( ( ( V_#TRv ( ( ( ( ( ( ( ( 5uy ɹY0=4%P@P@RFc`"\{>='v ( (* y9r=R@9 -lqU%P@P@P@P@P@P@P@P@EL1\o;G#׌ cK@P@P@\@VS~Qy@CuED?ttU` TP@P@.'lj.Xeb EN[{'sy>71@`;I$z( ( ( ( ( ( (*]^5~jFLl=>3y vi$dyO \D%D6 (wlۻ{s}%P@P@T}F N8pO{3-݅a<4i `cP˻N!J$è7j'}p<{O I] ;F'ȧ  \@ >v1`O-B,ř's(Z( ( ( ( ( ( ( {o5ܫ)bBރבAsg{`][J*8-!!l89=:P8gXM(tm t}@PqDZmv/`z3&$ ( h͝6:G'€!EB[v!T@x?f{<vcy h*"EUQP@Rɱ_Vz +j7s:VOrq6q@[[!Dgv=I4=P@P-W= ѫ@h ( ( ( ( ( ( ( ( ( ( ( ( ()jZ@P@P@P@P@P@P@P@P@P@P@P@P@P@WPEw0&#=O Z ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (+xn02A*#? &ӬA&[XV@J@81Ǧ8 .mu nzojou1 ćn݇Q;$@Yֶ\]ї4TSKmDėvmq9 lXQoSk`.1JβBQD%P═q%}BOOt\[At.!d!dP>=p: t,haUF@P@P@P@P@$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kdͨ$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kd,$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kd$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kd$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kdI$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kd$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kd$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kdf$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kd$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B ytOMT\kd$$$IfT     n ? uFDDDDDDDDDD t0    6((((34B ab(ytOMT71Ddp0   A?ØreportPicture 81http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.photo1.jpgR/'m{ Rbhe/F/'m{ RbheJFIFVT-COMPRESS v2.1 C    " ( #$&'&*-*%-"%&%  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzVTC-2.1 YUV ?(QEQEQEQEQEQEQEQEQEQEQEQEQFy((J>QEQEQEQEQEQEQEQEQEQEcEQҊ(=8Qh:QEPQEU=d\aݛk:@$1>~-&QFx5Fl2H-"m[Rխ5+X~w)嘟4(1'9(((((QEQEU-cY4/{,>ԞW/aqK X&$OJ`]?W x{Vj>T| w?𗉈<:a?ֿE(v}mҟ ?/Ǿo)B=? N=?Q#}p<\hJ4ό>QTx:Zt fC҇G@?T؟Oڂ^'ԴlMb̻]FFdKߥU_躂G9c[ې1xnxL-n?-Кi~0|K3n?"E(_09_{}?#|/>5[Ӯ{"eߍu Yw y?π;7|?(yi" $xb=gw7?8<|ek$ ThOQ~!P}('S[Ή3N)[&#v23gW;Fnj'ށ(/8v~T16}& v wWu.`cDHO ^EQEQEVԴ-cMTar+UqO)GKͦ9IdE !AQE|M.%<6ku_ˢE@4ߊRg|Uq_.taF5 U)?3O+ዽf{ O h`e޽M{"mPN*jQEp?Gu:k J+Ͼ8>'>o5+$ry| m+[<[[D!mE26rUQק]}{VH Ѣ)qk1?(:f xOAעEGqqD܀N<MRZD_+Pe|gk3ZQ^ZjzW``r=G=+Q]! dЊZ(=7VI,2)W`z;גx.vk DKfT!9]; ~k&W9}mf#hf\Oᕞ忳<șMj͎Ũi7 =e# A_m߄^EQ\o ։R V袊+3=IQqh+$s]?p:isOLcmcWk?ht {i^T9p׼Ycb0"<⿄P c>uOceq[ #/tvVwNL΢f21#g9caR6v=?aR =5v{B5Fг 8=jZ%R6FQQS ɇ~&K \\k$x 5sDl% 2U-neiEP̭ϩ>?qíi2? p3\O~ xoǷ~ úsHu ,xQ{=q #icXd+~Z٤Z tUN(>(>σ{ƃZGBEYoƿl_ʣˎ4Si o,@s"'%֧ycn̑Wx=(dR}(,[D)N'&ܟx+ ſMmDp!kuTs۱>ZO~K]cd^.C']ե $e'g^|1-/;|`5Y,Fkk34@ČAwPG@N*h]֝mo<qҸ?x< a:eHe ՛^? ;xUʵ´--+&r۵hZdeΧs*OtfyEzAVq/}AN>gF}Z-eK~(|{Mz:QEjm\w1 PhhW\ wC>W#Wŗ q{iRHB A\.a%k#kkC,G {W  !i3;ކ!p>|F5kե|:W00kt"F?=~*I)⟇ʓkc9*WHU}%ϖDJUR?'J5 @WZH$G1wS0GoZ=$ftR0ϵv|TRWXM'S⿇?]&?7Z>۝UG-`dj? )&-Tڀ?.Ob3vG<N1 OhCΛ|~Ν;[A>hq_w>MgS{'n @|D"x 9h뿢(@)wk EQEyē@P zQEpN>#̐یZ:|S袏QX/N㦹Co?Kkc ׎E =5CDᤐhE朹PԎEyO+袣纖8aAF =I=*γYwLID[352^ڼʗspଙO|J((+⇀9d2(_ğקּDQEQ^eG]ĞdYLϲD3nƦ?, d=:i‘ۏKg?QSNGK1<^#e?\xǐ%, ]9z}??9|##~9mǙLѰ2Hd r=t~X#JLiVq:7J> gUʌs7??ut3cE=R]|o+/bppS^ >n&xOgb ƶE;~.xIn(6~ f)0QolnqT$cԣ'_xkZ}A}Ht+SⷂA]v{ i<im*H~)x0zc-ndXJ$$c^}i k-Z/iq)G#,p+v/ m<2jQ,\Nm(#~I⟃]rg~ u??jx0uG)?QOdg~ua+SAf;S0X۔LkQEWXҮux7M4v[<~X?0;+s{$8c/+(X TtDz9Wc|+(+HG-=]QExgok(h-4څ]Pj*6h/m<ˉV5ϦI'9q6 =A|jHm8?b@5_~W.1> 訢+4|z_U<B+^ISn&bVq`/EφZӴ,YȂg"WG21r6.ry>>86Ht["DC;FNaԁzQEQEy?AO[lM,R';<x' P۴ >;Amc<۶3~ ?ޟ) Q hJ: >U5]^C$ⴵe啰AxGFoc׌ VG%#韮+3^PUGk[ƉpNNv^4/?Ch~"Jx=,! >^H`1#gRy]p>>3? ;o7H|33]Pc~ oXcȊM7)?>C_Q&+yl|Ou!GK'f|MUv(Qq=+d?X'Gx{yVjPAŽ?{7<DJdHG ?Q.|or?$?OQin$?~$$t&HM⋡qەY~QuMcE#||]>;h hWo=?~1ozfZwg~`tadT_xKL-紊FHIPp>ivxABa|W!qRƒ6 WZ+u??bE CUc I}DsUDZO3fvr~ ?J(5ﮭn5~rHpFWzcך]Z fhu &q#>aWZdžkFӯ*DCXlYEQEQE9/zFE$(Xha'|ٷdz}o*P_~t#h^C_ϟ|yt)9ۏ.#}9n$QI81J(+fOƟ(Dߺ词"P~=b msmlbҾ#=M?t8( /[69y>,ѳ;\O]/Qi:/۔}^ci2dl ' Jht/`xKо!_tR4/@>2SN?6Ӏ:FV/to[9^-A9Ï*߅tOK!}15E# Tl?\V1=8Q/>2'R.ğˁ(Y2VB>5#i?~h~* ^/[HZ|D4l9?u D{(W`iS aQ, Ң{ +jRNAOޱ2eM)/Qi/'Ǻq,Xۍ*cXzY4'RM8x?`|oZ|SvGW?O?œ') C~z(?ZֽQI'~ |0? J((zeh麭\Zλ]~ЎƼB x'GTo 0:v&xkQ.fdwﶞ0kG =P} QZ6 2<K\.?"t? 7QEW>O( ? y+WqEQ\ exw9?=ЍwW?OhїP 8MH}:=pc׍[H|JtYd>>6EF7 } /IF:Ry?BaxLa3\WMFᝪ{kQ޷u mV@O:\{F?F0 4axͼ:<HDw\=<1ggRfKy , |`H#<#SY|_A}G 1[>weAcwK6ohN_a+-Q{gl!9>I9R\oJѯ⍶pPʶDžZᗂWGO6r < ]Gὥ4Y{H݈0z]sZ|_M/+E-4q=}4w >WB"Nг4_ 3bixG[B_@% x_M GO C"h>] 4 h?/|Cue%ƏkkHD TG=džxz~j0nõY$B\ AnO5_IjI* wF72OZٴ(| ggc q8,y$~jھLeD%w$;qKOM(m,E gsnDѬ-Fҽ"1Ă(w$G~Y@-sMa?]QE@"@(;`t޴ЍwQE|OQ` S&i /"R i|؄~0AOq'_ .-t+"lHc=zWeRC\w%XVΊ((++ľ%ڋ1dP3$z"/rkG)%׉84 BnjN[o_Bx%\؋.Gmw%*/㊟K6_k7ĉrv+զHXQ˰_h;쎱(Pƛ7O:Xz=?ƚ|C}?ƓMu4o)5xWӯO謯.$1HV m#F}Q ~@zHK [9! Yoq4~ß,%SP1>ܬ %u=j*HErM4)N3'?{?iki?R\3Ƽz_&29żlI|\z[G|gQTETm<$v}gQe F|qSR74Rڟ<[w\چZO^xEEVOKe]H`fI?uw&7᛽CS26T9NC*;[EV]s--/&6#vs~w_Wm$?5L>(_Fl^|^o/1m;T?Vu(oX[Bۣ#]%QEQEQEQEQEUoD>D$p}T`{y]-㿴L#蚤0"B6G?J>(mpsS^&|7MM"XO=7Ct""LA= \W-7>.jv"F+Ɵ$O$n,r:q֪\꺜~4E6R)[ :sֻ1weJ>!p"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyza7 ?ǂTrTzRhP0ImZ( ( ( (24+OIwgh4*I;G':P4b(cH[EGP6R-Vf+j'xg@p'jc5h ( ( ( 7<#7EޘΖ ( ( ( (m/"Z|zc:Z7~?PY1ՠ ( ( ( ( [Av)Xޣ_EG@h"ڴPKB7YC3q:? ?SL)vTݻ33~Cy?;sgYFua}?Pbx4┩tUb2$l4 ( (iMWry^~v1.AۯӃ@DWF 2 Q@ (2~L{|F3g0ix3ݛse( ( ( (<16sP!7StV՚E][4_YVVuR͠,;.JC:Q@ Lvn&q>cS@P@P@[M:h"r~"0dae9zuP@sg$ٖG_fF3g*_JEN>';զ1zb4 Z7l?PS#?3@V?OSP-'='ѭ@h (STs O?9照Uo*35rZ~Ґoq$Pi:Hq'Ϩ=z6w5=I\~ր-@6R-nxG>o܋1-P@P@P@PǴ_HDͨZcЅK@u?Z( 9照Uo*Nu?i|7@': g\ּ)n3u>N}-69heI3?A@]{fi%G'_ tP@Rѿ c^".+&-He$p ?h(k+i.-[kwGR}M!]->X֘%/' h(·*\itM9&V"4( ( (%Ƶ5qdA.gF|dqu֠ (2 ZLHCm=?@%Ih ,C8zc:( ( ( (2 ZLHCm=?@ fo3t>iOt۲$ܧ:i}z㦀;O }SZBjc5 ( (%Ƶ5qdA.gF|dqu֠ (wL6oZC-#(YE@hp_5ųh;(N E];?1G^C4HEmCgw0iVGWpFw5ŭռ2e^96HeېlӁҀ,E%_Y,3 ۼlpO@*HHG-\>Sۜ [he]e{UT08yހ' ( ( (1?w.Qƌ.ٕ$.qz=5ė!.&H+ʦcrέ䂞u1@!}YI! @r0?6z&kd1H_P ?zc@+ hS⑳x|n ^9=Z ( ( ( ( ۏYOq>8VA'{4EU.W#': k5M䘾T1>lQܒZd\rh-Irnr'?E<_LMl| p ךwGuH h#vqÿo@,P@P@P@P@K5?}}]{gq@7 0p@ >qZ$GB^ pT|cn33wskHI i88;ݨi%EIqq%#()TU,\͌s(P@P@P@P|PL_:eg >^*# kq<29מ(,]XD~WFJ 3t\HN69-v_IuHLc;v3q,P@P@P@P@5dFGP 24Y,0dh#U#2q>Fh39DvSA# =y nm~5t>nrbG-v_IuHLc;v3q=o:+*aN #;ր}>9,ⵎI "|6@ ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( $$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd]$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd!$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kdz$$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd'$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd8+$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd.$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd1$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kdU5$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd8$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kd<$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kdr?$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)T$$If!v h555555555 5 #v#v #v :V  t0    6, 5 / 34B yt&)T\kdB$$IfT     s [C+DDDDDDDDDD t0    6((((34B ab(yt&)TDd<E  A?ÜreportPicture 87http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.example1.jpgR+ Pgh 6+JtzL tFF Pgh 6+JtzLJFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzk ?hbw/$gY().FGM&՝BЩ$):PvVZIo pV5 3Nx ( ( ( ( <`؇Pl@gx|B%$er|PP@k_}Z" +M/@h+ѿ c^"˴J/ ]׼?v ( ( ( (9͗4 :B$OCP@j=KhTK-=Cγ=l? t: l& GAL?ZB77yv9LL2ٿݷhqށODnӻ?ŏ#vݻր5 ( (:}\"BHhm.aZ̒{p}= O@P@r3|+f%s"khNߴEJP@RG.-/?ZǤ"+ѿ c^"˴J/ ]׼?v (8CZsnڔVE,G?5A Χ<-?STs ORh91A}2OJb29 h<7j\~H?ϿN(j (9͗4 :B$OCP@u?MoἸiǏ)ahh(͋ͅ,^f2:j}M! VL0:3Ei {dV[ f? €w5dk JN9!=kP@P@Pw,4y`!@7I(3G}YpGߓPI@P@r3!Ťl@:i uDG0)J( cz|][__HDW/PsE1h_x)(P@P@P@P@s^./%i@tIk֟(0 (}??Ttv9om=!\Ǽ@΍ [A] QjФ P@P@P@P@x>lܗa%Z_€ ( iL:$@N o]m੝'v2F'kJ n'@k*#]*(ª˷/@YCj$81> P@P@P@P@RL f2 #iQ ۹xdMC@cY{l'46͎SA-PP@f2ݥ6H062y!!;5y;cv6 \IH{!NvsZ5g|wnݜwKskO!ps{8_X ( ( ( ( k#H%B \s 61fɑ qp1:wi\/&p3w=#L_鬍,nqOnJ,o12-`Fqf)n&)U2A m_Ju渓ΑCmœ=>b ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (rDd<E  A?ÜreportPicture 88http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.example2.jpgR $ ' ݿ`  UF $ ' ݿ`JFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzk ?hbw/$gY()cI1<4mnV@ALu_ Ǥ:XZUqA2bBJJJBxȠ (c@P@P@P@PuZ\C C hO}3֟€+_6,6Vȷ%~{@P@k_}Z" ,h2i@4w4KY-{/"Z|zc:Z( ( ( ( /o :@_OF] (_Gmܟ{9ϖL?ZB.,+i@eMFiYVC{Z{); +m 9@0ɏohsbL :/yuvؾq1EP@P@WOkDHT18?-ݽ"kYXu=8硠  ( Svg|s۷t~P?ODn6#wXxߠ (+ZցZo*@\Ǽ@{ET-nxG>o܋1-P@q: ]B!)Y[8~jOxZ P Χ<-?STWQ]c%[rc%PeΟp' dsAxoYf1/~PP@RCV7] Zz'|AV.@P x> d8k@o"nk\>Uwd4d׶ d p})C{nF&Pm=?@%Ih ,C8zc:( ( چ5$(>pI?fh+5Sc={*( ( SPu;I%hePvahh(w ^$S{PP@qZ/_H  P=p*-/ hs?{_h ( ( ( (y(R?v (8kCZSiEqPr/tP@P@P@P@J_ Z׼tv)jZ@P@:wp 8:@7 ?U!ΒX1rGC=@ ޗ<$T iDm܃Ҁ[O*F' V:ed62cq.͜gI4n ( ( ( (+Cty)PmI7209 @P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P9XbK#^O m/.SKW"EhفJshquqOq#%ªa^X/Pg׸]elg)@t#Pi$G/n@  sm⽑7[HrQ\mg@ K}G6:&Ljz@tP@Uvh#h¨\|x LsYDW([k_q'#&ۅ8z}h/?oݻvs{-ͯ&Α<AH~b ( ( ( (+Z$i KǷ,= q:4mK0s&F 1k_!FIpA@173LX9=(ڵʎ9:PPY󥸚@Tmq}(ֿgO:Gy p( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (Dd E   A ?ÜreportPicture 89http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.example3.jpgRgGF.f!VuomCcF;GF.f!VuomJFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzk4 ?h. ;A-`Ic)8'?i~W(eޜZ7pH\gxȤ!B"@BҀ;-'Jl-Y, N_g?3K+y45TP}hP@P@P@P@eBMļ4H  ƑGb*Jc:( (<"0FLHIQQHCx=U,.U@\a)ޠ V?QegK@Eqw:7l?Sv)Xޣ_EG@h ( ( ( ி%!?4$3 ( (m/"Z<%W_3v0|9sݻON(R'"PcnA aɾɻ۴J`kYxVȢFgqz2oIa|džnF%{GF(f ( (+^çڵ"$*dր5,ggP@wiѻhha/;i/=33 ( (5%VPg7#:(c6Ywʗ1;yRG.-/?ZǤ"+hʀ%_u_@Jc7h (҈ ۹3P<5pab͉d4F' + 2TLr}^ ( ((xkJny??焟O*.YX4vQYlԓ(P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@n/^.:$jv\FNHGu2[ DSڴF~Ls@+HHG-\>Sۜ ogpڍ$%!F}Y8CW ?ͻ#N40ؠȁ]#BOėWh/baXl u9Ghφ7ȣk#hP@PY̷ip.&6*͌A}rtaw;DNݍpp:`%<y?mS8ր"O3|O.۷g8q@ky<}.1Ď1@( ( ( ( ( ( ac/eB=NqHs@ 5?a%(^+;vO,*H;n;,y'~1@ -Y& nہ~s|z~}ͰT7#2wt$r;Ck*%%Sq8c;vv[+JY  ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (٨#Dd@    A ?àreportPicture 94http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.mcp_neuron.jpg R6" 0xV˨F"TtF " 0xV˨FJFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz ?h ( ( ( ( ( (.nPӳݱ99 ǯS]d5,.?!>` G͎q@-\26Ҳ9$ϯ$w9ˆ1O9R;PK<1=~sF [=r8=88g~q;nݻou9 J48Գg [ ګ[Zᶀ z}2$iHsfܽqWX˫Yyc \Uy݃O_c V p^hiY@ր' ( ( ( ( ( ( (^]V?z Y~]sp]udZG ky]94@P@,Ϗ1'͏UFR)so=T{ԦkZ)hVe LqCs 4P@b!W> 6ے oe㪪#KKpil)<II׀ ( ( ( DvdDq.\P@P@P@P@P@P@P@P@VmJSזȽUPGM3cL_cL_cL_UYӚ@*; n+-{HmrYOߓ=yS[W+>k%FyC]3gLwcL_cL_cL_ -}wO@P@P@_OF] ( ( ( ( ( ( ( ( ( Vu żD\Ofݠ7Dd(V$-Ko<_&&l;ieZɹ P[j2ivɲbTn rHA Io pĻcBpP[3΀!t%Da(L(TQUz uP]Ddu 0 C@>-62Q,$' d&sz*cfԈȬ068b6s,]7I]}}Tc*Ms܀x\5\a08#<:uT}2ݤccv. ۝lqhY~S# ܡxRЃy-P@P@P@P@P@P@P@Ui%.)舱~jO@ .>qAK/"\R ]BnE"PwץOۏQtWs# =:}*18F91@ o$oGH7?0N&mL,LYH#'.>qAK/"\R췑ޓĬ9 71p[Σ$A; cwmN}0rj.cIXE!=(km!3O 0@,H?uB` 9"OS (+Z,&ܨ'>vuLoL?ȂP@P@P@P@P@P@P@P@5"3'h$rh-7*/s h*"EUQDI1YKಖôp=8Ǿ{P@P@ tY20+ = bba[|Ra rsMMXAmlCe!93п @a j( ( {." <}b&T9mȡdPP@P@P@P@P@!ԭ1i?x,$r8#hmJ٤ jq)_9;sҀڭDdc2ʥLq8x@PYvuh#d&6_\}(4,MB  . `yIoq7]ծn !P $ vfIc=8>'3Ey #Fkddq^H-H=LЍg8{|48=Z6g 2I9RRg哌9zПL۲4I;# K8$(1;,Iuqqvh+ S{~,yc}68r:PP@P@P@P@Ay_o_](z( ()Oj֎+G,@v1(PҢ 6Ьw+ns@Iht'H)p *~ =y)ɠ"7JB\a#8@9>@̾S 5*%P8Ar@=;Pec+i@÷`@W~Puմ)41̐E .{g?g?g?g?g?g?A}>ՠKh ,Wġc *Ji圁ԟ94v v1 nrT+d^4Y#F٢RN2 |jI4NI"㸒Ǒr{wٙ՚`8ǡ=1 Oxr:c*pGpGBG#$6LR!RY;w1j[;agiġU ( ( ( (#@Ul!wnII@P@P-Spu> O\s+1ץ] ( ( e뻻,!B:L? h ( ( ( ( ( ( ( (WW(1cߐO@P@P@P+f6 %'8Nrq@h ( (쮪@g*#q$3vQ@P@P@P@P@P@P@P@P@P@DT0X3r=x϶;PP@P@ pXe;!A47X[ܔKGA(_: @P@^bq xV 8_dw;3{vd@' ( ( ( ( ( ( (Z\GihsHis{>Aې=~U۶W[hR%vUpNO@P@P@P@P@P@P@P@P@P@P@W[ 5kuB-m'Ԏ@Vp$ hk 6[v§+m@[h0'aP k<'ןZKO-r8mH1FIEU)9zu?gm44𼩍 .F >hb4i$m]A4Ȭ͊rh}8 glVxO(?wt=(] KcY0ϯ4=ow6yeA\(*"EUQP@P@P@P@P8Ddp >   A ?ØreportPicture 108http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.multil.jpg R|7sI]l^$iX7FP7sI]l^$iJFIFVT-COMPRESS v2.1 C    " ( #$&'&*-*%-"%&%  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzVTC-2.1 YUV ? ((EDԑMpHRu8E>~*xEpQsO_a 2C"̨X2 (((((((((QE>Q\"ѣ 2?J@e$U_WOn6FP?J}QEQEQEQEQEQEQEQEQEQҊ(hWofQԭc͒4l̤d8?`rNqy>$xxoAw CrH| oLg9z-KI(!5DŽ~.wh5﭂' =Voc(KtU~$xFikv+m) DG4mx0?(((((((((((+ 3_OP$閶~KHKX.^b/=$A׮Wk/|Q؋KgJ"z$}}6)&c-wjslj9|Eƚݫx#O}wx"ii<7D|I~hھ {teovׄ<ɥN喣o%OaXjm2"2 EY((((((((((WmY(|muh\ C\}qۦ ^>güx8nºt? |A]V'ׅc XaVJj>6>VD'彵}cG5h'ѼE6ArWƭ#>?QZtT}ZAu]F"+q)]OfM2y}9OMKLq[P~|}Oi|7<UXӒ+е<=B& ,lHq_{KxY/$?sỈw5\V»KŝƧ{5S._,ut۟858>,xJݮ5tMŝ5k܄ogbol> 8 6dX[|=Ӯp/?/dgU?SPtt$fp?c\jH>xNqԮ T F |ij5:Aa!!%}Jv#.^iVSM!X O1}ql0te7ϟ3xĺ|Fxz HfX~!q]t+}Wƽ_HmW_|I&x_ZmEaxZ)I#aF=qzǀ.]+ok3eqS ,S&{65?sp}ckGY4lH Z(BK 2Mtџ kHXG 8&8ٌRsQ(|/Z,o|Wwƞ4:\?g?i͎2Y>wcI~{K8fc⦋mׄ/u="ņLX99 1Gֱ;3M6m^-ݎs)OR}.BzN^hVqqH̒#ʼ SA{id?KM(P<%x. +?Nӑ9ch޾:PP^Ho/\eߒkE),>+txecӬ așN{ǫZ)؛U q$r 5MKP,gsj5;ʹA8vcɺs{Y^$'tQKxYվ+#zx>!fqe%I%A[r y5vVE1.o|ACI9ψxʸ1[i+NX$_EǨkd<]çGҵ/>'HԭtxZDiuxi(Z, x,d#ؙ@?OU>1,`|UkχqjJo^#UZ>i{'wbb?Jz<KtkmL#DUUsM=ksbK-8ْS5߅9w%Ƣచu܏8p+~"(|?]  NOs_VJhm>MŵG${mlr28[QEQEQEWh7)߉RĞ!4Ɖ9ș Fu[= C"ƞ#܅[D}$RJDFG[4J>Qbmq\vYҵKgpX6[>*dk)APӘ>vA9<=yj7u5ŃHm RbGQϯ&M6zZpFMFk6@8R8rEni'|}μyݫ>6Gw?"ФzzV *;X/-^AC+BZS֭é/ΞHbs{*~ϊvu{y|9 kYnxqRkt]6v!m4j]&/E4?O}N=_?֫ic`vfw 襲OU.q. k9[5%//uui:.eYAel#SW(((((/ .r~u?j^\is"ʏpJK>EýwS!^GYNka^DmgS5i> vX[E[؀&>GVߊ^ &a׭a80fW ;Kitd0͏Q凩z߆ob:bhƞ\8r랼CXW:~rMѠau=WFAq1<ˤF}A0 sɤҼ9vuxēE$ bQ^ƻk[dv:o,S)f9۱j ->-v5l4!$/ ew:ẙ"BbsN:$5鰱t!*?g'@"F?б< tpDXTB"Q-s kemVi|ޘo:< ? Hx32^["\S2b/|qX|׺k ~G̣]#ou ȥ9u*5qQdi\xIVoB}lnΫ[|$%U&tv3Eio >T{%dMxe.틋l11mh?~Iuș{23[ sk[Sֵ W_EQEQEQEQEQ\o\Fj,\< ^ ʿʭ_[BfKxsXk:$z{;6-rWi[ϥC8/ݢuj!}M"?0u~4}o-Cn&k\`zīQo>x;OMRi_ge/PG{Cwv>:nXJ%t,ܗ?+X񿉼vJV ,hnfF(RV$^2/ľջ.yK#Fv}O"(((((Q 9 dp/R񆡯-|>Xf۽^Qk_Pa\桦yxrqf-ʶ0 XL{GE4S[{N|"?Z4֭%ҵyd&ԤVyu-Yi6NxDӼt|C2i6vPDE^ m7Z־_45TY%8G`8ܟc^y? = דaק,hVHv"k:^9}B0(鞵\@57W#~{sNȃxė O٭DZ#a"-Z7-ZkdOZUC[״io}CinH,}O. B9'x2'm#]WJl9$k;噿y&k?Wxkl.4B?xm#%ˊċ a4'vQA[{v9/>JElDk5QkyDyc㟈ֿTFk>C/YͨI*v''=+[/ē_K|m6@c^l=|;1OxK̓jkYƣգ_EtvxV=3št#J#AWpEK7gaoypzͮw{Ǭ6ЛXII?fi[ɪOOVU8~%Y'ز-Vq7I3sᾓejJ5 ve Hp}mEGú<0xZH'#H#st+ɤ |AF  %|`WGwž1<-ydϳdHJ'6=kh'Kt}?J܋{Qp>X V_ k7j,:E|B2IQ9< $#s^dZ6ta\DkFrz5%Po֙&kMj>IkXnu h wz"/Sl Ӭ^9XZMEcq??48OY5q;G+onbҷÏ0f?ծ򒢸mR(ZI5Uˊ]5kkjw.Bmȥ 3uKkB,n?7s+jxuvkxES[ǰ>4|-?S5.dr9k:rr~_xV=M؏%ke1$Nڹ|= e#9NMs<>a^ׁ h^/’A臗zHi*Ij<%^+[5[tFw'ӎ?Z߈^-"Ys\~&m<3+r쎃j\k:ɢIea =xdlw8i6?gkoxLa5qO𣑼z`+ʼUk o{%#($)Ĉ;+<ڔ>4}F{Y-R;U:νo^}i"FN#>x^$.^q'ԊO¶v|Muwmen瘞/5M/eJс 5Ĵ@{yaakiGom(@)/-2ϩ]kH3\ݾ4b|U43Iasq*Zoau][\lko[xwH[!)?S5EQEQEQEQEQEQEV^ci&;KTi]rs^?k/^ 'xWN {~|W׮eùP;\A'.zz\SG,|/y7wM ;,rrȣc?yCNOaҴU~yd  d\6-Kķi? Y=_ӊl<-?5,{ ެW:&u&Boi>ttp'q lBx_ yN]^;fCbb0 KˉDW1)2|7ܥnb!*#恣Cᎍ]]$&ZKwz o_Gud̐54WYN6f\?c!>x_2{y(*NzW5x\զidmӕF N h~[%o]$j5xfQ~-걶Gc|7Zy^%=ΕzުFLQJ<948ٮ{>kҷ FJbnye`7č'sۜlO~tV>'?Zѭ-XbF#I?;wdY\^8薖OЄP7KI];^*?eX.uq4~ ӣ^|U0OPIiu}]4q_)y$<pY sں?wWĺŭ%a' ⽪htO8 μ~<'w[kڍJ ʂtb܎gcOh[7frp9w``p((((((((B6't?]#G_5&VXL f cj!'m#:;Nf!:k̽p]^yh\v~ntQg"}nOx[?n+ "Hi{;c 8R+oLo|EZxVlo Ţgo^).V)X;m:杽`{:j!w'[u1!MC70 ->g6MA15eMvcknR֛Xf ;qX('z}xM ݾq~f q{;ś>!-h>sO= G'mHDe@x?Bu6"|jYi$V-n Y?r|0+LY]"jT)h(((((((:mo@// 3Rr#ҴZDvtSskDGj< m> isF>W#i6i3,/+7ܱ~GsYڅͫhZzn|z?z9MGs,?lٿw¾Ժ;3jºzf؛cQ<}Mu?ix$4#ф_ԚOMO]CR7:֢?R*S+( QEQEQEQEQEQEQEQEQETR]Bf(ќ jh/c+ǎмU[> M1smc4 Wt{S61vܵ;vڷ((((((((((F|C[}OV&-:lNpH>3|G}kh?|]^HUgf$`/g? ?)Gȹyg'b((((2 !eA!sKEQEQE嵌kۈm+Qts m*Kr?B)R;hZF 2I8@ QEQEQEP ҠOBZ[4`S2?ЭI " -QEQEd:WKve(pzW;x>M Vt0. SIӭmiRP3[j6Uu-c/gOy#&2vF <ֱuO%6rȲp6'9MռFYZ5YlD{AeAmo-.yzm2sW18mouZI#xBj^5k}m/T~[ʐH!] sh注iv}Z[hܡ+ȑpr89U ((~,& s{92 A?tgf|CC o}J ) .; ǭ2{WP\*ZjsDG݃Ǩ?fxKH % 4/"BP>R֖GZŧèByc|L1~Pԣ-nPKxŲƑjžUG6rz^:QEQEQEA~˅s ~I|B/,ђV׭M7Q6uekq![Kg*]m5MIfJ8,/<;aiqiq-*sA8+м1xo3i;]vJ0$,袊(*Ei7,ء>S/4{,^f)|2v玘'ұ.w1G^jh6-c8`sԌR|8Яoے]:E>Py Ϸ9?JYeV2, N.0;I -E?Qw$[w_R}E X"'V2xKmvЭԪgv[-ӌ*폅toh7pF++ն2NM]4:4 [hϔĨTQEQEfkvWk9ٕ%u(8RsK=_=fJ0YpFG\c5V o V((((((((((((((((((((((((((((((((((+>Dd^    A ?ÚreportPicture 109http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.neural2.jpg R=jW˕le=F]=jW˕lJFIF,,LEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw ?h ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( ( (QlثKP4ՍYvP(h ( ( ( ( ( JZ((P@P@P@P@kyy@Yß@>@K"IJ: JSp\ P ( [uZD_zK@¥ @ӭ.&O?eq&~4ۆ/ !=(f ( ( ( ( JZ( ((h ( ( ( (  gb:x]4؆52 GX[$w€4h (:PW#K㸑ƀ,]^{ d& 9P9(.^$Ee?jXAє΀@V[GF*X@,ku>iKn@BD.P0JM2r2(h9O7?\X20h]ON{ 'iARHQ<"9q@xX%s,@mJK5"?^>nZNms0<um.v'2MNq"ހ;*ݛ2ijO-X #<u da8se]Na s7#h-\vr=h#րFSY\PuƶfpY@ PNFNx#] ( (+WingN;%h[+ J` RӣyIU4"m9'3܎~ J[!͙(,lY['s%D=#LiQ(x-a;x'1mӯQBq@m5KK$uc(,-$OeShm-Fր6`6e9̖ѓ@Hؙts>oNZ2Ⱦ\$gZ,Ks+,}H{x\ x}ܮq@P:}ƹc2ahH"an_UGhsy#Ǥmd]I*ɩZX۫g@tf21o1<4Ԧ6.:c/M,9 pAio.#2.e</cV0Hf(^hx: yM`No.nԏTR0Z4t mQpHȠ SdımqҀ::(9VU@RK 9Y7~իtYG=($m#@˥F2Kћ|*Ӳ>vw6q 9h>n!*(ao3~mZ'^2z6 ƼL׮B؝9lv.>c@ 4_`vҀ$LXrh Z( (Mҭuus2v Z,4/.19#5tΙo(*@PN><:ccV hh k21)GYq (-`@DPP@R, }GJCbsejV2Mu\ZL0:@m]h\mj( J(;|Z$uAEN yZ yKX߷fb5ۓ!P#׶;68[Q@h ( (+j3OaPڀ!!tu=J?4btɭ ;Vmb8@P@S7:]Kd}G?Ҁ!!,rɔ?[@P֤u8w)[4nR$T ( (V҆_PȞ5,tlJS-ĐMNJ4V9 !%$ ( ( ( ( ZR9mr_j_e)w&^Xw5#M3'wߘejvr}O6.):S@1anSq(P@P;F+99Udf @u4;*Ic (A3i97".>Jҍq :0 fю"/@WC z(≵+( JG`~/@P@94[4:8ݑ ZJZ(3@P@P@ n}YAq.Q}b@W0*8ڀ!KKʰ z ( tHCր CP[Q$diM_HPZBJGMMϚ@P@<m?8 [-65Hb\G&,?Od+(##cQ (mecxJ!)n2I!Y {)d8!4 R{ #(4x,$9CAo]K~^3Qgրn5>1iXq\\@$mj*Pa?nnd>@&oˤX'KX1¡bPz(I@ @P@z( (h$I(kx 3q@ գkNǟ(6Q3j "eM}q@%Fң6:Ր&tcg@P@u;Sya,K)Pt߷ZaU;d^; Ri93֫oi͟tQ@PvΚc (36 ]")8?:^[h< Mk,s&:4׸9V'ۢɠ h9A$G_P/Ah-q@5 $\ vbl6(&l#HLw8#Pۧ .&@NPռ~MqqBB$/Sn($3f-隙I]fm"R@:;Y{ W]VM ^ ( ( ZJZ(HR@r6ͮE+F"?!~n@P@f_i5ڬff#;?ր u,1!:4llˈuchzР35KiWsH~8 y\y}(Za` o;I$F;\:.~%ڌ%6@3-mN _[8 ua45Ưg| s@L]*zzl_3n{`--HUlAz_ӭ~ٻ͕'BhԲ\x;_p9?tP%8c/+Qԓ@2$ھc@l+ s.0h-P@P@0Ƹ8WK$oH;qր.@d2ZL2Nr(W;@Njy̠Χ@?e֩sum+DWpї=N;^=@*8n<PZZ( ( ON4 C`@1I}}h9m-e%wOOj ( -F&((rZwh @ @0FE!E$J\ 4(ľ"068*܃OPjGNPF㣠GK̾m'v J((4gcQ}"&dFր@P@P@\w2FJ(CGa44:\-A5 bv @empX9aH Uޟ$[f~cހ-ZCvT(P@P@)kBcR@P@P@P@P@KXnF@>@ Ѡ< @a>@y€4SK4K6ހ3S}@]C}FGCڀ;~W4n𿑠:~?#@e9Yg4zK}CJkwM1 f MG@P@P@P@R'o660:HZ/a 2} h+ ( ԿN?&4h4gnzG33H(P@P@P@!`$=Ķv8g+ԯJONkY+chcL.r aO0h-<9(xkm}s@o"[f6+~ )o2O4H 4QT}>>/m?PxG| Ƞ qxN1'?1nke6R*08P"is3_Ȏģ'q@P@P@PP@djؤ&{c ,/#ے(ɖ3i><~TjiJXz( Ն%oKA L+@?Ic졄DT=P@P@PP3^[ؚ9ť8-'qNw^1hoN<$hN86"*ɠ^4r36QOjَ P+灥PDlIcomlҀ&K-R$|8hb*gho4k=ƀ-Av## 9RgG`@=@ĄGVﴶt˟JѠ ( ('hƧioJEOۯnmf(7(=R!VP @ 0vP@2X3#"2Ջ1)Fen(Ŧch-:3@ @P@P@P@P@P@Uo4 Cn#_qP;>)͌S/_ր!mő 1ӧ;P:ȁу)AP_7 eϨ4xFG+!h;[uMz'S"T7?@P@PY+Ks']Ҁ+Qk)on?@ gh.~ϣ@":ހ+/x6(2G&ȗ?2p l] (|gny~m#O78ٞs@( ( ( JZJ(h ( ($u ( (*a 4.=]:Pt=("7ki=O*~*ܓϟcLր6BjpCP 4 \U׎}M"툘m& Ǒ#wURMZNf#]ϴXWLyNow܏f +kP\s@@u+OXMo =. $7Vs+"?Üg;G:= ZE&#2?kP@P@P@P@P@P@PMJ,-7;"h>-*ڍ̊[q{ky=Ÿ?$Q@ 4g*"% ( (3П%6#M<+js:6RG ?_MpG2Ecܽ͗,\Ftdg k/5=V%4ѭG2drhqG #( (+5 2@ED @P@P@P@P@#2q =JIqlHT PZ=FRB\H8PSײ`Tjy3*@ZR]BLcP@P@A{tV ( ( ( ( ( ( (LкvRt/QRt_~q,ZX}\\"Cqךϯ@#}?Pq+5˰F(GOERsX\T,%Pd ?8#@m;-ŸӠ ( ( N،+\tMh# NvOŢ!=k- ѕ(6R7h,e}I ZUmv37 OH] 2Ė7nitOCK_I@U zZ($($M&8FG=hA z@ @P@P@P@P@P@6D&H(+Ah7@(2?e_*S{RASyPEZ\Mq`vך`.1 nUP@P@PuS]ctx4@ (挤OPErZYՖ!08$ @-%O~X (9`e,jtc;'r€ /m@@mfV>,P7pXO A9l@ L&dѹtRG H짊!:('߈FWi@P@P@P@P@P@P%'C# h)tB͊ݯ%@JyVmJa;/AG@5 _n 4Οy4f ( ( Ԋ5ո00>vhU-#P,VUy"h3=ť]׎rA=s@P@PP{>!] `K+&@ŷYy#,Zctmۘ3R('[mE>Γ]aԺmSk,n0$:͊갫ÈmA&~c@P@P@P@PP@P@P@P@b#hP@P3- %nT?9rR9nvgN{Qksh&ζX >e4@P@P@y@.t{ nkZ;c:bٟ˛u61C  ( ( ( ( ( ( ( +lhڳJP4VLI3O+]/oKԑҀk%: @PJ)urc#j#%ݥ͜5u$gʀ"SGi3M甕.0H j( ( ( (9ae+,jaCD۬c/@ 7Z{eA8?Og;lv?\`A" y㶅@Fb &@uEiXGU=EZ J( Z( ( (3Kb;%i@8-nZ+ "*42fX*]zF^(]6{(]€5( (( ( ( ( (+Y[] O ?tI/HZuWN8H*2P7vROx6KndF>ހ7#"@U(Z S]yʼZַ.mhUEK@Phx<|9 4P@P@P@OKq-:I+gPJZ( ( ( ( ( ( (ʮ0iA6:|34!Mx@ũKyH{ 5H~If0jv +y,uT:r$^tHH zu=TfZ3KV'uDx͂?ܑ@m,6O_€0t9!,#7y9uP@P@P3Al34S@#0JC@Ӌq#JEm#@P@P@P@P@P@P@P@5QK34FmoOE(,wWvO»y[? [GrGޱh?i\{xw; o3$TM]hmZKe/P@7boߴn40zʗKd"ѕP-P@P@u{ٗ,khs3M3' nHyy4Щh$/^ Ux5t@08 ( ( ( ( J(h (*]6|K  ˽MGZˎ!m2}<4γ^$iqPqz,EA$I[b1mHz(ZeuzڐP@P@seK.lX,6hHRqq3V/P(!wopzjcp' u}$?o+FG,uoA@ 1.ԃT@P@P@P@P@kwVCRLT\PX늫-|JUgO@jEk$AL׷H& V(;!G_SȤYbYX> Jw2hտݳݖmg"4r (H0((@P@P%O2'LH@:~s-7Xg[Ҟ-F<9(aG@l?:Cu ӼNNv+]u=SfFI,Fs,6 FL@FpNZ3nZ}.+m4V84aOݴa"j~k8(?/W\?_h̳ܵ/3@]jſߪ@.b.cEn 7Iaz@ @P@ I?P<f\ƥq_?mLtC+W_yoPg'al#@ í'%4kq@ ӾΙ/hջ)qƙ/@ RN#ӂ[-Ǩ'$sw</Z֖!U?:@P@P@P@P@P@P@%eϢep,€ }D5t?j@ @P@P@P@P@P@P@P@P@P@%-P@P@P@ٿ(L3(N]P9b ( ( rόjN=h)h6􃎽>\v@a!I4j=7:@'ހ'{dEvʒz)RdSЃ@ ( ( (+j?c(߷@VuK{* ܸ?*PF.Kcu4"&W?{?Lf%[Gtom!Gmq@MɍN#T..u#" ^Ҁ7( ( ( (9PEwDIil,$MArLpq9Ѯuh)ֺ]ͭ8<l@P@P@P@ nۑsEs?C(oy4rHbX{K M20p(uUFсzP ( (ۻo|P0m㱠*q((*2(J0@Ҁ (h ( ( ( >-]&Q6[  ynkc4XҀ!)nebLP[@P@P@P@r:tz5mFO8 ).yv3@Xcw7(? |+4}#,L,Ԣ F:p@hKv}lZNx K-=KPi |x|tn=hJ ( ( (+}I{qO :mV\3xǡO7C}dDxZ]-""9;tЭ⵸$sؠ -~;mE?ÞZР ( (b9.) 4&W.JM"4@el\4KGW `r<14[D9ʧ+6U~'}L= WM%W +nT- *( ( ( (Y( HoyIs`#$ УEQP#FAdV##: cP@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@Pt$$If!vh55#v#v:V  t6,534Byt fDd   A ?ØreportPicture 110http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif bCFa\6oh2FQnCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`Dd  A ?ØreportPicture 111http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif bCFa\6oh2FnCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`Dd  A ?ØreportPicture 112http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gifbCFa\6oh2FQnCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`(Ddp   A?àreportPicture 118http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.perceptron.jpgRB',s1 4s'F',s1 4sJFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz ?h ( ( ( ( ( ( ( ( ( ( ( ( ( d17J}P@P@TWgv;{}^@ 'ˏ'@ oAzȦ=vhP@P@ ycW=2q@04P@P@P@P@P@P@!8=(%[Hoƀ5 (6_cJ V豍}LlNhO@\fKIm`?wR) R`9zPg]I"L= _W'( ((%"_,ʼC+!BF~z#@֑-!Z4U$w  ( ( ( ( ( ͹&kaߩ 7# `€,@R{hm+--m=Dh< $dƀ%5O ӻʶP&e4Oʊ{T-5g4QcDTDە F k2ӎY4ءh2F'rM9s^݁6y9@ڊeXϔ%wg@o{ݧ^/?^&~/3~{b.YJ8P[6#5@3#Ki{;'So?_P^N͎H TvP{idiE7^8a@ ( ( ( ( ( 'V2?vހ-Gi Ž4?/(%P2}B BQԟS@P@nYC0=zHbZ備{^(G_R}M6_b LY)po hP@Pq ݙ v"n]ǭK@4"Q۷1C 2Aހ ( ( ( ( ( ki;p:Y@`9ՏX UЅO@;j3arĮ?] QPPS#E4IcWG?€ E֧$_ VYb(ݻ?€.OK\P?P?gM,KepvI J@P@$s*6}JkfFWld6O?*נe Xd(7M>)4z~@P@P@P@P@cFI var>sHhF NY1$nf#Y^ ~'^4FUzڀ@%U?j78{Tk[2эҹ c@%p%G}I΀,P@Rg$dPNTڀ.@P@P@eb}VPP VɮY!;na;oJo-a# n P@P@P@P@P+?dA~'PZ6d$Owjү4Pt #jmOa@ * ;P6hՔjmtz ( jp#eO=N ( (+mTNOG_Z@/JÑwN ( ( ( (+_aj:/v>:'XLf tPMGOQ͐nuVZpb-P@QZv V֑G@#b@T ;PRp4Z@e3zj ( $ DwJN~iy@4P@P@P]5v$%?g!mf' A=auo-P@Ce~NiP@%-P@WM]yQp@ bLr[f~r ( ( ( ( ˿ )%i.Iq=;<-5pʁUPtc?@ e7?N9܇mSl x4x0=4N{y}?u>J}ZbIOZ-P@p?Տ[  (6p@TP@P@P]:1[p TP@QG'@ ( J2=E&P@VԿu\_A4Vܒg#ր6t(,SDVcjiVڈ2MM+,Ld@P-Jy"yPsKI!P4 {rNʃʞ0s|sҀ'9Wpʀ&;-g+$Ϛ8?X k˰P}M0B:ʃ };ʏ ~k?B+ydS#\@@P@WOS'hlހ-P@ HO.=3O761 d84f3.ZZP9,IҀd@ϝנMq X@vᕱ#V8 {Xu-AOQ(1eέR$Čq?>=1blq2A}7ɿ7N9ӭ2_(MwjJIXl# $($&)ا,y샧ɠ ϧc>d3# }̃gPnn#]<3T!ԑY ɒ5NT=hf&֬G;րWS[KPG!-!DOjel7'<(iC>T@e+@EpZ#+(--Zrsƀ}{f>*pzPz\i@|hόY=I?Oq wh̃L qYEkfOͧLd؁u~@YQ yԜv_3@ g>gYwOʀXZw}˼_ȠOmiyƩ8 s@4"Bm1Pr96Pm!VM]C)x"!_(R ·r11@ݧ7JA@(42,FT9*K]qO(Rŕ[VAwP@j S>\( WS@F*G,QעՓ{+sUmFIZWcuWr=eTS8c dCp?3@u)l;'lNs;?(qH5 -@PPb!$kGjr?ϻ4n ( (9< _W Ҁ,P@P]kF(8㰠 ,6O/3~KMӡa LNm%€6w(!'# "rWt{_(ZϨ%`q@[_ ?M7~8\t[Sգ;kθ M,0-CD TP@\J??ʀ!Nt_ʀ$Z/>d>9 6O|/H??x(V \Ep8 0 {P~s:+"e8R o)x~gjʟ@ G gQd` ș!3G>X (l6F#ր"=?@[y@C/P}__,%y;Cc#Y ( #=($ a0$GtP@#0U, @t/ ݰܶE JKdWh()K$@MMuH>aH4 7}4~(% .,_1x(qsi¼c$/R]>Bah\ ACa?qMd}Ӓdm7Sh)ʃ@ #1ʁ1cYo3D|$~}f RxVH(J( ( -|Gs=zP( ( JL;c@hh)dN!3@B(U`JuFq4r*ƻ8#QE"C:m@\~ mۣ+s9v)lRpwB# bl?€$,{YN:}?*hsfFTQ@UزB`}(K4[GMsn6րoYbH(H#V+j!3ր4c"$jT`@y@{V lPG1!@Ct򝱯>€ehp_c@ x}Ā{?X ( i,F ;:Р ( (*3= .(P@d䷌y?] ( ( ( k 9Γk\]`XCH͹{I00 ~# dHci$`$UI!1(5ETQrKihbff\?A=Uo& rb'G@ (3f&\' *( ( (7C3?(*iw9͔/ր-@P@P@P@!@_(g2>c#.?=:vۘmчZ@P@fSU~ݔƀ4 ( ( Vj\|v (Y x(yEb ( ( ( (( f)K%}d8hJ-G9yV4b15[YuF͏Q@P$D2'@@FAMI%\|z~Tz ( ˾}0ǒ#G<4@P@P@R|@h33uCOs#P@P@P@P@Pt?_ lhdbms}z-UT4"efc&䇹|DzA@ qI3,{@P]G@U>aP@Jyn'Qq4v ( ( ZhIG:r]"-y8D4WHIA @P@P@P@P@4OFfq@ 4grTdڀ*Xefq$Sv".ahǸ >ez Ͱ0>ur$+ (P@C,;_O#€,@VHbC"298}O.{~W7/wm1\46cq@P@P@BiP} hh&KR[{@P@P@P@P@ΖCK)F1 :}Z$UK~[(OG *Z((XDNJc(Oӵ а'#@`I:}K@n[eP;ß΀,#2A: "\ ~,|3ӎj(?P( ( [O2v*SXԠO_͙sj3`76q4s0cWm^3s<)4v ( ( ( ( Ȋٮ;u8fy4NPK4Yݲgh'?@quia@0xuKBpϤW9cua> ( ٬yӿրr6?n@) Nago!x/~4nf9c}e'%-ŔOLs{滆KYR3J*ieO[](-b\%_1?F1@ N8ePZkyR@ L~@7 Bc.BGzw#Ot{"8Q(~@RrFB~7Wt8 a@ (+ڋȄLSp.Ҁ%UUh*mٌd( ( ( ( (1-!Im-1 }=hhP@P@P@WQV6D.UѰO x{\ߺOq[@ >w_$ז&Vf^X>@P@P8Ikrc wW$-hiW0coҢ Tr?ր4h ( ( ( ( ( ( ( ( ( ( (bBRq@׶O{{VοPjR5Vm @P@P@P@P@P@P@P@P@P@P@P@hh ( ( ( (X2R# P@P@P@P@P@P@P@P@P@P@P@P@0FHTqI@P@P@P@P@P@P@P@P@P@P@P@P@ـDd  A ?ØreportPicture 152http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gifbCFa\6oh2F@nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`ZDd  A?ÜreportPicture 153http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.greenbal.gifb5x_uo xz X?pcΐ(Y!\NX2OS=_`]xB-fKKF]s()e?t_x5 {x6hP,"V*RV F2+~ ~ҿfˇ@#(Ol~/keVaѴ##jҬmtvC2rvk@~/xǚs=(Bqwեѻ,nT(5xWIZ+= Q.^;HaX NA|/ZG|eMWIGsX^)-HZH_k<=o[S^.\ʬ2xyW |7k^5-䵎sSā-܆osV<#|?ux^{ɩ c4Sr(/=MY˧x=/4=j dZO*{4 䟖S,`"~NEڳB]WB<+jW:5O[#&yI6e!r[54?nKi-k]DGhnXy06.89<|i վVt&;uve"`I'FHC^.>x xt[zF"-0p_h>m猊ٿ|'iT%mB[h۵ɳlHAXTEIWr Z~ h>,uWmJՙ d#Uqj~#4Ir63"F>nvq33cCH$_tCzw|A^];Ecgl9 F%|#Amrt[/ ܼȜ@ֱb+۽MúXK/ey | i!"HΜxQ\[ c g?Yx? |?-~&J0-mNnFƉ !6>,'2΂+ DG;Q!y"AhnS/,<#YشsM6e #FWFLh~@~o A톣w=ӵUjJmy'QfKvj]{DcXJ-`}>dHFeBHϖ1€ :Wxg6zN -q 1,W,{XP rBq<[c=X6Z[ZP+V{t$K4S OOsxMѭ.u[O6{y4,pȿcQWi +Qxz߆|Awaj![K,nXz]"];_~x[z]ŗWm>5Kma*#C=T0E*Ѽ%῅^"=+Hk E#$by&vpH>|7%̗6z:$VO4mm:mH?Q2N1vuLRYb}2KĤ7y7*N6r[VfGԴxya2[g"X8;+X~?.(_x<4FiYuϙ\f7iccqҽ:-;YfuVDʑ5.[lg<=k:l~j xtkiV :!F?^l>UeR׮5Ӿ*Ͽr8x~\zO j{L<]m⻍BG)< 61lu_xkxMaXhM?ſ KӪk0_% ]Bhe4l6.&!5K?x-t ]$eq;HI>UuឡixoIҴ{[8`$ëD7R2A4돀IӴuS{"^[j%C%I(;dr眢VUTR-QSJI-"ԜC?f߱^ prIEQEQE>Wk:~▊(; 3zKԖH4ͼbDI?0XdA{{^l쮤gLP27H`pqZzŸ ^?ěƽg˔ɹKyeԎx#>O)|>J_8mղkbLL Y}LeJ(ˊٷ%zV6ɤ-ċ;K2]bd1ܫ3|%|Gx Oeib;LD1DHh$dh|WVzfxY=WP BH"sJw5-5o=tV:yt۫绽6>Z80L$66P*rZh~#Hw֣'F&UWܻg Wj?<]u7 k>*4-3R,jon`XYK PH€rk9C-|]KmB^Hk[[;Ci%Iv*YIbG>+V-^7Q5[K/k i1JvͣH&ݵH9w|]7>KX5]4xf_b/pK>]Oxt/Cj8yL)Vܴ{p#"ḳcc ZũGn{đ((((((((((+3Ğ4?i6̹X*}F|;hma٧a#} Uu]/OWrFUAժ o<2R,<o⏀<xĶ}%W;h8bVpT3?]oŚ&Kh$y1vd%H#j$մ$Uv.@mݔ|Q3|+Ѽ/3H>$k:JJEYVS$yV5o]ս6-2ʊ#x>brŰ9j6%Qufx# J M`z-VmfDUa=7ۗdiF3sr@#QEWῇfg5_"H$v) `2xJo~CyutRmmH\6W$$xTdd#ҫjz[%ޫ{2\En+`%b>#R­QEQEQU5mRF"1m,幔^sUM#žץ>\܄03 kQ^A'} |P ,Qfˮim /r#8!k_N#-ݰ> @źZ km")auQdE"ȷN:2?a?|AkKOxgpPO3bK:0>]M#H6|Tש]u}vcIӥ[GϷD3i7SEQEr+sŵF50u=M'eR,#:,$cQtz~ַ^xX[ B]Q"efP]r&1F89{i^$ͧzM &"g䷼[:A}+A+6HO>%uMl#nc6#r(lA#tmgMc{ss}.{ +tmL4o..ĮBۜ _ˍ'RӼ^;Y,QյPGI2^DB:EcMvN~5.> >lLܲ`"(+| FeKi#4a@up{q:Ok{x[Ɨ)^ǰ։%Rcu39@D 3$ 5A,=Ɂ 33̤d;?>(h^'h\YČ ̟C! $[wz||Co3LawKK(-|IPTrʂvn亼"I\*$+`/|ZԾ}i^){dd\eo.тHnEQEVVO ZAԢR ЂFAֲMe'x2}J RI.%g-9 U|ں-P$.=X>-QEQEQEQEQEQEQER;hTdp~?ZխbPAܞBX1b(XMR*vM%Vw{,_˼'_Rq[tQEQEQEA}acYͧjvP])I0"GֹsESaZU-p{8h;v@}H/D f|Q*>u}:M5]MӴ [ٌem#S]Niڵ_iWז B X(+Ǟ׭Zq(b@auSD$<iq$?dZOP m>r}jxͤH5 c<!ۥH%aۻ]qLϿ=8Vdɬʤ3)"l~.Wz<)Oʘ c@pQb#VԶ:F\TPnč o$ Uq֕|\:_fKW8 `FFUAQEQE^Pb7m<>=?Dufݵ{!-xgx;hTdp([Wk=N]fMX]"Cks!?dr e$W=.cV$DDEP:; uQEQPXjk8.b'>\ч\ `Ɨ>Q泍K$,?"|_i~,]IUk66z*tsMQmӣ{o<-|ޜ*][AdZjp&d\xBk&Ald82s$exPO]VO/@ђc;U | 9!}F3ٲx95UMs1,N;A 1[BC4PtS袊((>=?>°ZyړY5ΎCheD:'{m3o|5B{ge4QHI-&k|OhvĖw!K!9AF]ɷ?.Nſ[\ ADpH5~& ,[k:֛~ҬŴhs4{6 UI%QEVV_ "Q,T) F2:BטJeu="??tחzfZQvrFU.Z55[~-hq+.3.T +);]F@m &yNH!奜]Ye8QEQEQEQXu':.!+U\ bPAYpWk\񿋼gZWKEjGfrd@Ax趬;B`qde B&?1:O<7ܵZ{$~T7\LV˿."u6 E( BD# ^)4ЏgSҒaz3s9Ƨ q^S$3,I⺪/ ^\^.12C5#`E`v`*A",AiR4^̀?˼wM`35iqaxCs DǮ;5:_R&(NU6Ӯ|M A YwE^G5K|+ǦkF]Md,#2Yz5&'iW;[nUBzr<^9 ahw^ fyei ;1qnGyOkã=t_^q|E~8[B_"VI}JS޾\|y{8><8_>-P yxz C5QEQE6X3$U  jSvma=oȪ>ZHZ ð p#܅i/Be,۟za793=$Q ?SoQ/7=9>jOMJ<18>#.-,p;a Sk>/|3Go >8Zh+Nڔ7Zvc bڲF2pJQxŞ1k_UΟI 4rP2`UtUkMOOih4+z#cсVj=:}CP6y"D2xI'xukښi=ܤG:ӕpа?>%x׏&xON^q;}wSW&V&T*H@$)Omm;gXxQ+k"%dAwH c']'B<=uc35$]@\ eeF qԨ(+> ~%C87>s."5@G -z] !xK|Iq#E5PJNR(l2<{>J^xwŷ2GE %H0D'cy<7}j+{MwL^ՕLQEX t2Ԇ~I]KrH殓E~ TҵŚ鱌}3#zPT ]Ā xW᷆7=hޗCo{zK3Le.8Al c:6ZhBsi?ںHw9ewӚM;1F[kz]-;'%$M6IVy<ω_3N'$ 1&,ʤ]J.S |t>)>ռT߶lEi3=̝171U035<'G?f_#u>![5j*XEy #kl>'dO4H]-|^vhA+63]u?MpϞn;O7MDlт2A||7N{Wx1?L֍M+ZKb kKq!7qO^Cg׋2āfSb-@n;m][?V0i!ElQ^O|;>_u$Z+Mh `]Bm=LcX(]9/[} ~ {"_9A# GuQEQE>|[_}c෈|;tHmwo-̈́ʯo`.Ýȹ=kA'?t{ϑqvH2p}J+|gmIFgԼ6m^#Ԏ.dDulgYt 6KognH9$ =+OǟhGO7h>m]˺&YB4uR3`~EQEp7N35(x2yͳV 16 #=x'5EQEW ]#⧊>x^P~k'#JKlg9שQEQ\޳iZS*?bsu 2F3PxVC'1o}AbGwE_h6YpD"\1Ny>KB>|wա%q>Ww+|_~ bǟ_-/]84mjQ|y w-#oK۔_(l. L͸Э'*Y|!>VfNT|&/h~*Ѯ5H`m{Y{&F)¶9NHwR7r$q*԰rrqӭt4QEQE|Ff$|pDDR5 =N۷_ O:ͽƷsַnC'aepsʻ%|om=sK6QD_lʈh$>ՃocO}7wGßE&'>.oe8dq(_>&GVtجiS︿,E0O 'x;k((^gu=kG '5 k%yC2M4< n*x"隤ڱhV_l NV oA[Be5%I3xR>R;RGSOZn@Ώ"=AvwB/hcA๝g $Nw]\-F{[ZY[kgoD6q UQTQ\į~}_.fnb&&z2Q_ 'O"vgX! qGy# >9nğSi}[sJ{aFh0#2{((((8C" ɭ|I񅝃 :7^]zs}xQ܊@ > Mj_ZMj hv &pUr݃sC,|hMğ[ʸz+ qRRxa < 侽_HYŚ'͈4}X05 , 6947[ X$)&첈_w(*x2𗈷g*A"UwHYO߷OK;V5$e}JEu" `>2 9 [̈́٭ S 7kEk/x+ ^Par [F>RAo/~| ~1~_:VMvA >QװQEQLnbh.!Ic~A `lNyXVriSD^{2e֬bG`<>qrȵoxJ+Xך]@V]>_U?ԵCu[mOo<9ǟPTkb(@P $ue0T(WQwU  8HR:TpGA<گ{HQӵ) MV)L3e7 /"AG#|so Ssp==fW+y f9#9gПx,O/G>mŠx_JAӯ'ggtjPxv;r|W*o- W ݢB Ż0?QupFb r}* iu;da`1:Oj_9t<_E r*~Q?0nYa+k,r4V||E𕏘q5W`tχ^? ?حUO7AԚ袊(iwĞI/洴 ,V\wGpJ㧄!izƁnA.MBH"G˅rI s^Y Y} xnoa$^k('jm[Eѵ3kMj1 WUY/iZySQcAv%H/6qASkдGM 1*DO,F}8=jQE6X'xHR PGq\>-N6>FII㟼q+Oak;ZNwe?,I8]Jml ̶-🈴Gu=_R򮕧[)eިr2sf_Cе]J/l{fc.#˩F ``!"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzv ?h6T€HH8u@Jd3Gq& Oyd?#8ϥ-P@P@P@P@C-Q8#;UK}$wqb6W@TY\@k~ހ(F3J!4( ( ( (3dM>Y.2# c ڀ-ZEh@Q@m/cmmyX_l䓫E sr.϶(\ͻ^ɳt8đ]^J ASJۈ!B`ьvr   )u=h卡 $٠IeMb~s4gNQ)*'@P@ X;N7PKKXw(x_pNp8@X9dY7mOJтmw((P^>QN:P։!\ޙ (]FlP!%~\He 0E EU.:PM-P@Pv1($@ÀH#8@,ⶶGd|(,í*u U|!كa|POcGƮ $yvO؅QTԚE6jYPzzf,XI7s ҙ|X@Dx]I93ր/P@P@P@P@BGP" 9+pZ22xU̬J"iH6ov@x<qu-J$|2[ ">@ 21$MMM,*9 n(`l+F'@P@P@2WDFBh+7cw?Zui y Ҁ+Z=?-O3@[-ı4@6z Io/.=phZ"Q4Gq€3|Bh&KHz&,EMwaIP+ 22@㱠 LyC\1?QYy+Td"̾xXF ʩ9_N&9"ry{ǯ4Iwtɘ α0}0:P>gt$1ƀ-P@0ƆA!Q (o o3duinbHPNlH'$s(e4r1'),-ern_ZA {PhΪ7>yIFcw|P ( ( ( ` P hYv~>-*XaX$zs@ AbBa"Ps>ӭ 9+?.01) F3@ Iqq#E( O> ( ( B<iH|{Picog+w1!~(e8Uy%ɠ͜.ѴbP!R9< ( ( ( ( ( ( ( ( og h ( ( ( ( ( ( ( ( ( ( ( ( ( ( (I@P@P@P@P@P@P@P@P@P@P@P@P@P@P@P@G/(J( ( ( ( ( ( ( ( ( mf>1X ( ( h7GF>,9ldz (ݜ7e:-#a8lX ( ( ('@ ( ( ( (-5'Juf?uN@΁l-p>g@rYI@@^ivg hڂ08nWUc>[fZ7l& ( ( ()I 08 '@qooc@( ( ( ( (IP ( (w ya^ ??^ IbOנJ+8IZh@/Y33@P uvco^7{09R3)?ط tLۙր4cu.΀& ( ( (ʅeLdP ( ( ( c( ( (zI@P@%rzƁ$ ( ( ( ( ( ( ( ( ( (6;X)ݲA~kI q\%=WCO Lq4@ @PP ( (WhP@P@P@P@P@P@P@P@P@P@xR;MrP61M@~xgSf&g'e@P@? ( ( (b4( ( ( ( ( ( ( ( ( ( 5˪Nggq{&xEcXC^H] htU-GSUqʀ,@?@ ( ( aX> ( ( (8iW- fPP@P@P@P@2YG$* -lD(/ c9]%$ (0KX68WϧjXcnĠc>q r8QFmeq;mghS@[+e*|΀-1\~T$w0 yh^a cdo (t^q }T[v94%Mp>cs@5԰AjQ2|ώ4K}T( (@!a$e8z<< һp{P (( ?5-2K9cGPꊃGe 0hYaz@YYp&O?@@P@P@P@2FAJh$,t(U}6(_6Fb]}N2?XPI TP@5-Z 9>sC~o_9>Y1 @&B8r{q@o3>J ($\9 :@ $ǘqz #gq\ICEۚ[yV&4rY{@1{j ( c}hP7vwnHV:=؀W;G*88픒Ki1!3m^-5#|>U|c__a r7i<8F\ڀ'( P@d`Ix秵M@PIA :Aő@RIΖ[{Ec,)8_N(ə_#f ( (EDWf8QxPFc \=ԒK]DhHP-l;'R]F]HV&6hՔlN:XX03 4\Fø : (094P@%zx~tb}:ctbw!@} <Bh@ (7veC}>m?b8mVIppsڀ*P}F"ҭ |ǚӔv?(z}BT9&/gP{{l0%Ww@5.$`dL1 GiPWaZ(#O*im8+N/:N2c8QJn)p̘V^LJ”':]؄,[lP5_ |(<("1pP}/ _"/-L_*VzcP0A@vdQÍhj( ( (}>@P@)B!fQ򷯱R}݈{inmxfHU.EJr`a  cw͜}q@@A@!2ͼQ>5P@P@P@P@P@EqPf"h,ƥc@@#Yxԯڀ&GWP(PS\$<@ HO~c ( ( (Y* ( ( Jd\0;HO@ [F)r?C@UPs@20vWmfn}4$p ? ( ( ( ( ( (Oķm%$c?J Cv 8tvPG ڀ",YwFb2~!OA8 aKBIDԟ ( ( ( (d_@P@P@P@0FEW:}9<`1u96aP@P@P@P@P@P@P!X-E=$= q$)6RE__1ܼyudzPCaOܧ  ( ( (?O@P@P@P@P@P@P@P@P@P@ \cte4B#%#}=ޟkz" GCЏƀ(3?n zm>DAXc Ph!uK@P@P@P@ _@ ( ( ( ( ( ( ( ( ( IIÂ?ƀK8'ǩ ( (!>r A  -P@P@P@ߏ(J( ( ( ( ( ( ( ( ( J[lDCMiEP ( J"P<=P@P@P@?T%P@P@P@P@P@P@P@P@P@ Ҁ+n7'wa@B(U`@ ( ( C@ 9u=hZ( ( (J ( ( ( ( ( ( ( ( (9eHWs$ rO3O@(h ( (.G8\eXzu4,(Uft4( (8x% ( ( J7HA~=(wFD {;QPSc?vL# =h ( ( ( ( 7HuzH`*l4P541ޒ4Z-Z) ۏtWn J ( ( EY-ZR izm54;v,-;c{ր {X!b2qnfw3}SI$.r}֠ <'\Y&;FFZٮ"I >nhԳ^f\>r٥h΁$K@P@>x$*Tl`[hc@,`: hqEyPm98@Mqz}hJ( ( r%[0RB@wFBhzhnw4&v(Vޱ#یPFf[e&ҒZseqkeu,`@XHvDw9c!gM_1O9s@{[KI#SY8>jhUw20ǽI@P@p I@G4K4/tpA D6@]%`G98=mE&ͱʆ`?r*j:|>dMrמ"ϔJuztk@PieOJ, \F\LBP@P*m(,8d|yxab#RhJZ( ( (&+k } -P@P@P@PtT!"w'5tޞel ' tϭ\H&PN1N(0՝ eu8`M'tf cy$s'v?7Hqcvٰ9ȠWJ\)ϔS(ĉ#4ϥĐ&$9 ( ( ( (_VHqx|z9@.D`(D^b{֠ SY%8*# I(זo'ր\a#@!H@@P@P@ wXлu4h-J ?$g@ @P@P@P@PP@P@P@P@P@P@ eWR4Yi"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz$W ?h Nm16>O04Zo,ֶאI[ v tHDR[IQ"YLAyҀ4( o+ɓm;ns }/O.mhqx$vdm98f^jW/vb0ؗ?ٌր/Hcv] ^x  t%o^[d@qz9 _kH*7MtA CʉCrpGzt whxrW qX;f8@J@y(XdՖ{{[r2bg-܄~rp@7l۠Ucz?ۛ3s4N*F = {9l渹{ HLmրfffD%@`qt ր.W7[Acڀ06, ?9c۠X~(f82I#*dpzKgeQw 1Uߓ/UJpx0ifbO*Kc{22~\sLg1PK|I>ހ+h D"(Є m ֠,aAEB*P Ӽ)22!pO89$8b-P@0o8ע(J( Hbg>:@P@P\Dd " A?ÚreportPicture 7http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bluedot.gif!b꩓oJLwLmn꩓oJLwLmPNG  IHDR(PLTEƌ{{{lcJJGB1431%!000 ssckZZJJW tRNS="bKGDH cmPPJCmp0712HsrIDATWmD0a2Wi*(Iu-2˺d L"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz)S ?h (F` !7viwb X8X ( (('86C#08 v"/P@T{c#**wa(*}ϵw-,/4+ r Ԝ# -G6iM܍]}:K<϶dz랎ޟ9'WeuӀpZ<q7 m憎>q@/.. R[^+ Y 2Lg`Ox 2IWd}ۚ`qʌ+@TX839Frp;1#k[H]% \,0-ҀJTqPAARM7XAy8 ңaM#%ICu<rhz( |ՉRbT{ކij6\*HH?Oʀio#+ځT>8$ = O.$T('-NKNzwԠ"߱p]1I': ( ( (Dd $ A?ÔreportPicture 9http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.equ3.jpg#Ra{dgP%cF[a{dgP%JFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz,y ?h U1Il|t@ ( ( ((;$ A ,#b@F%Hd_Yw`7 o[i/n{{)"qalJd $JW=w =8 ӥ{(dngQ \pB{ TP@U]لB{z{`U(u|yI#FJ^K/^7c jP~`dlgB P:E4Y%xQUK.W;sҀ&`[yW;/升@Cs,"slC, gn?w}EÆ]3Z Y{bH1M?Q{+FH|Cm!R=H@᥺!*g$F8#ׯPO$;|Ysa@w*d5\F(̏{winciVĚ[sD-[:7Vһ•Gx\VHD{ s oz`8$/'U8nX>3 lr<Р`[YmܐF+cj[-kF1zt'yei4M<4J$6/68A 5( ( 7{elKpĞ@^iXgy/b#q` OO TP@P@P@P@1$*F$@-=J( ( 0Dd % A?ÔreportPicture 10http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.equ4.jpg$R%h^Ό*sC(F%h^Ό*sC(JFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz/ ?h (2d/V;4" )#q# ,neX6 Ԩ!ds€-P@P@P@P@HOn`XSgvqpw`d9#DnXd $䍼Fg'@  `@$lNU{Gsր,P@P@A ֭&/ PWj&@' #L-RU0gw==+3٥7Go6$ g18Z֠ f s>X~bI |'n 0?EۏQt}w_]Koq,U |MTԮ-^(<\a/<^I${U 2nāt$>RR{QoOʀO4@.lZ{[tBF }@ kyHR4rvNݸ g:ր&Dy']c9=cҀ!73,HP"Fpi*9R A P]4gv 2S@-/K"Y_zO,rw`TF YT-f…<:sCҀ*ln%y.71#<j'$a(2 =ƀ' 7\w)yE! #EEr6|2*20V b8Ui@0o8ע(J(3PT%BV_%ۡ TP@P@ٵ Dd & A?ÔreportPicture 11http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.equ5.jpg%RO "">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz1 ?h (+xHƽY?@&j 2O-@KcL_ h"+FVSGP@/Vԣx́p@#s=P@P@P@P@P@RbomEYGR]J\ynh,-"h,oR54XJ[x AFb2~ ( WSmf#̻یd}y@2,.u _ pwi[0E9VRuPA Io pĻcBp (v-#4'fr'<=~]K~ TH͹Us (W4|1QG\`ՆQ4BEʜA_ƀ$ ( -䱨T +8w P@AǛ,2&T{PG8O@zǓ@jF"ryW=TZq4:,Dc@=kPsJdpGdB+(8 g:̀b(c} &) |@=H, O9 T&w$ʀ,;FsdMW\= H*Z ۼH621p9@ǤZF#U쌩D7H#/PGdv&ݦ~WݎӜ tͿUΚ 6`s۞h;q -!D_)`۱@[1.P30(Sa-a"&ϝQ7 pG_ }.kZy.Z5Hp8\럮(<_:!$M?&?4(.l)dR \u[\GuJG̥NA #h$RO.YFU.GҀOvynB$Tvu+DP[#tTI?4ූ}3lBAoL={Ps(bs`s@;c c6~.?!DЉ%(IbXM$k#TB#rfwKhq+̭(Q xCF)I#n?,\-aN=8ͦJeI,k6.F#8Ǡ#NsY]M3m1cl4zK G~aڀ&?i~W([{ ;W/ok.F GSN(i7fF(mT; Dux (5 {{HM`D1;F7`l@J[m5͖$u R?dNq"k3F2YȊ4~Uns(@ #-㴸%eo6b!Y7q3]c m(we[$`GC> /xĻ.?SgْcW2r Ż€-u𵕫qIX[Q 6okiXJՋ`Ln?֗}1@]-o# *Pd/h?@( dVU/C-3@:BAWsbPHɴ2:㍈ $dǨݱxyZQ$7pN:  Р ,uX UbPHlB89.%1E;eo@lP2/nTGbz4%RŤp Y@3gЀg65+f-Ј+6Es?x>R/yiѾu8n99g4M~r ;Gc ( qh2$c9¹ PG@(,d<8$(>€h%Hw8}X td݆X2ɠR 9Üz P@P@R)<"J1`w1.>P{Et2I&4@w$Pul+CdAҀ&$8b]ơTg8`PP@ul+CdAҀ&$8b]ơTg8`PP@P@P@P@P@P@P@P@P@P@P@P@R Dd{  ' A?ÔreportPicture 12http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.equ6.jpg&R sX"n)il $a F sX"n)il $aJFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz1 ?h ("M(NHF}9SUuDgcUI's@ =2>K $mё%PU[f]۶^3{Ny@P@P@P@5cFw`,p4]yDh`R{Pao<-PEhnq,( 18j (זe3!w@~3:$Orm dp#*G^XV`ɐ9q,H>%)^݈m|䅧1Vˁ4lIwv7᲎Q[a$*sԨF 0gAv h (2+'P$ԍ2D9`{1IЮ\gVCCۿ@弚-D5Q ɕРڿ/r쁌PWہu9RI$ znn.,vqy qYKdUUqF A&m,"'Ys8$rpI#'X, O9 T&w$ʀ-PG`Gd)tJ4n+eo-гo;Yyp~;[3Ew/ڥ\hB*yJRHU0wx98)p93@)i;Tv#~]wy&u=˅ q s 9 "+>"3:ԅ;m?&!.Lt!x gsӎF@.[Yڹ{{X!r0Z8œzq@P!` 0ZC<)ip 3 id0/]RRK6@1ր2%Ѓ<Ǖ͕N}=F@4-l˗F,NS{C#g[8=p: t[Yڹ{{X!r0Z8œzq@Ey żP:$NI)BK=wD h#Y*lgo\ې ϲ!XȪ7#M-~A c0?=>:wYdff͸`: TP]Ddu 0 C@q L898GLր-P@nW6־rB؀xnegH4~ 8 ,y?6wvQy6yc V W ͌Ƿ@ (eD$_ܩ hJ(...m嘢@SzG&9XrnS:SA$+5L RC.W^4oq"` #4-j@E"-jb$cpA7' v\Z6,8|rPl##Qy .!TmI;@99玝2v ( hxFIaUʀ!~A @ V#z*(~$ ( (;VF{<B{.[1.P30(J($(K "8?292tz3X$(2m.yWUUxWz5( ( ( ( ( ( ( ( ("Dd ( A?ÔreportPicture 13http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.equ7.jpg'R s_JBu>:w F s_JBu>:wJFIFLEAD Technologies Inc. V1.01 4"@.0&4LCPOKCIHT_yfTYr[HIijr}Qey>"">WIW  }!1AQa"q2#BR$3br %&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyzw!1AQaq"2B #3Rbr $4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz1 ?h (*>j 2'h""Itd`AEI@P@4Vf#z~& MWO՝Ve$A[ uz-Xng^7:FOr2j ( ( ( (*]ylFPH#2 @;O#bDw,ЫA@fO/ 0ru@: ea ܙSwăNhZcon򈤛`,~eBOicu,y#!>Ƞ $Y$XÝc1 )c7C"Ĭ.%K <7n47,%f-H 1?xPWQ b@F}'R^H#|[iY6G\q$xа6+]|$c R@Pvťq@/.ikiy A 4n (2+'P$ԍ2D9`{1{iD6u-@WY!6yNO2:a+d=Ue1\FR(0tWFAAP2Y[pgt,<^x+z_yops#w|Р _ 7Kú[H EXF 0?<t>]%7;ssj:L=GP@5l+CPG6FAO>W92@Aary R3 n'~T4hpO=GZ\򐘀F~oFx>(IԔ-QFpA݂#9 2vO&+ ҈.Q%ฌI2`Z-ZmǙqg)k8dﱱn9k,iX%v@4XC;c!p gG~ҥgxXP)@yq%[l[F<:;3m?(JضH& 4B?A Q^hCqoJPOe?}mon" !ef\Fm};{i֙mkcPsހ'ӧ{5Afl|ہ8:cZ&ln%2eo$՞%$$Pm፡YPWJVtRd L2)` 21dg~N9ogmiû (i7fF("ƊQFT`(Gxۺ˰Q2N9'~t^m6S,VHYROESXn!usarp:'r}:Pem# 1<0y+X|#RJ.`P8נ&ln%2eo$՞%$$P[-27d'c@Il#@&:+:?#",!aXPJ6Q,6VȽ"PGKq:۠fE}p2g8=LK 7 \*ˆ N [I$;Ǧۈ3@ug?%DRʖl#:0ޣ;pqo.ƒVu7l1;UIvېhB h:Be(0`(99_e.s$md3?P -;IϱCy3<&1,2 {;P[1.P30(΋"2:FeaG }Z}&ve@@ (!3c ! 0#7]AlKm!#<րmF@cS@@/ Żͻ; A@-@R)<"J1`w1.>Py-[3#(dCcןOn[KbsF<`hPKx'\&2q' АT?>P@P@P@P@P@P@P@P@P@ـDd ) A ?ØreportPicture 229http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif(bCFa\6oh2F.nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`Dd * A ?ØreportPicture 230http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif)bCFa\6oh2F0nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`Dd + A ?ØreportPicture 231http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif*bCFa\6oh2F3nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`Dd , A ?ØreportPicture 232http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif+bCFa\6oh2F5nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`Dd - A ?ØreportPicture 233http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif,bCFa\6oh2F8nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`Dd& . A ?ØreportPicture 234http://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif"@@-bCFa\6oh2F:nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`dDd 0 c A ?Øreporthttp://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.bullet.gif"@@.bCFa\6oh2F=nCFa\6oh2FPNG  IHDR OPLTEą~1,StRNS@fbKGDH cmPPJCmp0712OmHIDATc`F%% *PIH)TH"ьJA@ J`s& BiF (IENDB`ZDd} JR / C .AAL-YARMOUKbZYȵo:G!e6Ys?n.Yȵo:G!ePNG  IHDR&sRGBXIDATx^ ]UuH Ix;Tр~@V@pE+"* +u $x$7s>{"rX\1o9ƘcεSO=;'fh=cg7Ni;>,9OǼҚ XnXع@W&qW"2 MDu\4&9^y#:+}0lo4j+!&?^q@T\ƣ>OO?0ްf J+r4n+n;60ȫxu @>7HϨ0nP?M^+n41d… _?k׮袯|͏<uޚ}E!e 3.8ܾ_@Gh$[$086l~ot<+]+: }9Ag?uOԼ ( "W_w+qæn~zwm=G}* uJaRt8/CW|3HG[Z1zb-6nH;erL ,.Ÿgcǎ}@O_bҥN;rʾ1όٍ1*+.bi'JXOϮ7'w=\jtn;w޿ۿ;3a„umذ5kƌ3L:Ǎ3M"PAu#OI\KsKկ~cwu5<|͹jYzSO= oxlLH,[l̙ %4Jإj$\Q01S'OO4iŊ+ @T=C'm|[vLvyJQST?@k#ea.^xm%@0p'#LBY%/yk^׿okT㒊:b^Z6/ >{ .tހ@L>A{rG`[wȡyY%@3JToyhW]unZgφ\ =wl/z'Ȃ \x |4;u{]H$J<D;).#?(2f#߸ J#RIN^QGR'7WIt}gƌ[C}K&i[oo[! !yŊ348 nD~pO~2]4{7"ozӛpN(\w㰍@Wr Ne/)pDQM(6찀1/r( w{QOGg ?CGQɗGJl 7pWdB B?moyÊs%K!3Qb?ϣR !# aфB9P7`2 .vP{`0 L1Y(/Y䥎 }wmT ,Jy3ϟO~IOFC_F-MDXWor0W2Z*G~_GiX|y ##ۡxQš>юkzg1As 6$O" ]$jhߕHícCygd \#4DnN; өq+U&TmֳTvZq*P3:>L:bo0?&d>#pQGF\?Z2!Au_iOv:]Ia 6}kwsPB_" TTҪEQwZ"mF->=CdҜ Mmڳ0 ww@Eo5\nT ,,^̵^K2]b_BȅΡB~:ѢVwPoFks"2y-),u- ֘!/F}xҘWND5&HQIűz~gyDE,^WUە="Tv ոs1R@>ƎFfw\tI<7, Cji䥠܏ό}@ 0D@)!c@!Lǘ6sRwdFۈrFƻo򗿼t6K<@DCb(O/bz6gV>6r%4NPu`|;30}/5 9GdGc?h"Q9:(O}SxP6DEFȬ "j +=eQVZt1oS^bjF4(|Y?X|Ej׾555YvhV~¿@|GdV"-Ї>DbtZx68JDc$^wa"3k?!Q \5˃>Xͱ"FMu =ijqc9,X΅dqDGD5RQO`I. ̩  *pEi!jGoW:Ca9眃Lg ݳ}RhZ<`$#Ah`d\ _u˿<  yvv3b'ZIJ6@@xw~3V7.ΨVxGB CԗudX3m- q_3YJ 4EW$2 )j؝#g=W(Jzu$Gçԧ@ȅ)TA2'1mSpB 2v݊'\2b9FhCBMXH u+3R8UưQ3Od8o{&;wnmLIp::JIѷѪ&N tSP:2ƣ[r)Hw]'p]w// _({ xt{b1DŌqC<|oUaQ%0[U6 5@EX+aԑ!kAc1o< E]$4hǰӖ@;88@"Q5g͚UaQY~Ԝ8!ӟ4$B4Vudh6Qś ,x\=x=\jt.t|-z5M$ "1 .Vo\..`XHH*w 7*+ ~a}^$N++5I$cq'AMz@L&tT*Z`U ,ڽCҗ@b)Yt/~ԴQON&=CfwNK 30N@,f#uQY3D0PqH90I}O ? 2qq!_%q%qXxKд֯ i+Y} {Ch@!.V =O&1"b$ ĤJQbzOјG2G8uJER] Cl˧ !"V&|TpY!D(o(%AfqQ__1\[U5"!C4. b%\JHo~ dͬ77U[)U fn2+P?~1aA寸r :T4kÝF~F!gAt0 M.Ҹɴ洙gPyVb+_ K hEurD) *5Q,h4,D#(~j_/b?#MO|PI *4d4a=7>Bf;mp{ 7a_M7Ğ$ho=R":K1=?Lri@Vi5:qt 0>u|;UDAEe%\]! ed -j畞!CSISt\9: P삍ˍAJ4?S42O4hQW(B5čA/_v${~7*@-D/5Fqyd3dfΘ>e tr2ר- So28awS\ACёVsB1ȥ:ҲmuEUR8 N^9|ikDPz!CԤ?a|5ݓ:k, `~r=B0M֣ %MEdNdƩ N-p*KeJD gTPsN GhJ_|cG<ӟT:<]q  EWŽ ЬK PCȌ#W#(v6CDo~e^ԪgȠtGA=+ӏ?2m;- A,Ȥ}` O%,ob3qYgrƅOb tES2C  tjƑ7b|٣ p_(3hjʼnm^N3C,vGôbrb ~Zd RajS d-`viYYUjl PTeWA(`n@ϐp|WrAEoGk 79lP`?&tW\'F%a뱓_HiF}Ҁ]&<4"λRgL Y_;U{EDEbmhޤzG%ϙ mJ&pfp_(%jO5z r5E n :wD6 ,fL1L&3ZmVr2G:̦l*v/ 20[|v)~D6m=C\-8lj׃ DRYB !TW8XX;JdC;*nh+u4fKЉ%!7%sX*#b(Q…?Bh+*R1iAժTWpjj2ك 6lTrhc4MSs~ qK6cW8P#Cѡ$ BiT;y{7+(1؍r@>nbBĨM!4AĆb@jN6 vd$Ѐ: P P&A|@#ԋo}[Q kPFLiQT2 0pc`kIp4"1OЅnMCz(K`by,."*IIIzDaGOr3XÛsL,^$yE[dDb!,Ef΅AH)'&XM)0 @jX:3ZSOzFes"U0-AʌQ63Z)Qklgs{2JmoeBF:sB4z$hd2{D5}ɐ:#avKutD-q'(tUE\)V-]%$o#Ք%A ~IuYPgCCXӚk6r%#-P?5iEUT:FFI:p.6)iӛV]M)초f E͵K~ ~9D?Q~,Ra=ET &k]p8NuƌcRe-VsnC(6=Oy$:ip6OÔ->]é̞ixa%=@xg]'U@3niBVYA9"&5w4?l`)kw[ x뤑O.k}!-zp"C JÄ| CR]km-Ł6tM5e^ᅦ ԩ{K˄GAu'k0wO}"(EE4 euܐAjb#%A;S>*Mz6M95RH%iJwYgۿyiLs3L` r ^oطǂ3bڴ b?Jz.&v d0_bJ #`\fV`-B%zsDГР) "Ƒ;Ă 5 ^l4+ u &O2|l;qLܝ(A h S8JfT3GJԟpHWnp,>. !=r=:h~ǭ9?2i0D3w9w_9/h> \-$$T%=C܂_ݙyVPs3`(WR.,Mց*2B`$ vn12B+ᘚ!ӵK蝒؜5M)1{> VF<\Gv.0 dMSe}M[#Gߥ>r} #ٍ@YȁZ{liml{֩5C*"#ϔVbzg/!}a@#qs2y$ NL3y1kbYآ'0oZE,݃`V3A8LnbԶBh5eu+YOd=զw/J&@͙č `-@gc-L C: vӀ"/i- 0ޕ)Dz67\{M=lPEӞٽ2lN0Vp#;r#"TE 2wA?RkC#DE8?~:v;A3 jSwmXJܪ68͈} CMvXXMۖi"/ȵU1bb[VІH?'u![#]d$ r¦WMZX,4M!{Xl [^P )b#7Hdxu6)CGRMI/uN P 2bcY S'nzpAGr}h]nHɈ"hJ%Y=?:FF I]7 ^ i @`Z\m MBP-˖ M>P CSyi ێae#C0"3k( IN#fO*||R-ej|f*c(O"£;W[}P_|w).j,D OR9( 䴕j*\X߫l+"CTS%$*Xl2JnZv<na ["'4t]?Sd5˯Y6NivJb⒛-im쳹*`aŴZV~az+LqxTy%wb ΍- BjNfKzd@2XS]ok@/ ?4sk\8;W 쯶P!*!Q0yC4.=,|5R'Kp zS"\Op(Mp7c#AX::H{P霃cgX9!f:[xk}Wc ^[l/Db1I>8N$N0ܢ1sRUF$HM6|7']#: 2^- !QG,g  ͧwes"&-\ `ʾ:+;0efs䩦^!W]{糷?Ɠ@B%{?sSlEΊ1'92㖮Xqo$j,ju&0oA, 2VK̨0]mj(d3P0G-%يivȚw\(Ǯ#' sΩpni掟{?ri^UVKP(&\13 ֛4%EdFi1G=CFxgCT_N wo[Ib""=(uKr $St9o`lMpu67{}TA!.8 r]'Ѽ["#S>^!6Xa rHf:C8&GH&(0R-i.UPajJ$0p,8mGOpTΰ_"C92){UR5'Bz'nZ88lq$o*UB]]{id_߲iCapS67DI|vgqU kXkh{_PD țqEUv=@9Q{+0LD)g2X(=CKj_>!i\̙}%ذ`!s&: ޚq5?"ͮ9sW_d3BJC4xK#g3@Fʍo~=ܗ]K`G Ӟ<}; 62tjWKK"|ᅺ.3M OWi ʢic2J1 ,gRtSf]5wFe&|G x߳5}4aXXiF m TA_)+$GԮ~iH|y-dvGO{h3DMqBv  Q@Ӟ!z v|H*΀ä;: 큭O~6s4pf{] hWh樧Ԭ !NF*}k#I艐LĮ)ƙb7$#!!Ʉy7ݣ%kЁ%xLŮ8?b."FkBF+rB.ڴokԩU;S9??ϡgȐwf5boUQ*1a%/Cvxi-=QBܠ,<]eAW0qasr> M=<* -|x՘ҥ>1";FFn@F M$E!¢lNu"ц NzC Z\EPE ̙ȇ>u\n[R2ٔb{M [6mH`숝HVgZҬ+(:IӘZK?{Cn#1e,c;ƱEVcܝG1*ı%6I)cun΀h2$0 oA^,[..}RYrcd42d M73D8D_ Zς< lZF1gDUYO"0\EPÅK)M,M8 +;bbn?W*92n9hh [SKlÄX6':t^wM C;r/[;2 2eMz'PړTƴ~%5l/Ўt@n.|pr"5OTOŘ&>m'85 󫺫 | =ǹ2n-[%#M)ԯWc^`ݚ5k)9XВ[l/K ODfc۽iИ7W64Au()Xh!̴2X,TNM | ]7nAHd0ik=/5( }%KbR9m8[c{>VH GHz92i|x{9t 5' Uc.u3$l2UB4"1)!]gh7dp僃6m 뿟W$:/g 5n;LW3yIM#dfƾafh:UaYgQ[A&UQB$]۬Mdrɀ]M [TTQ^E{6Qlv`|r* qUڣ/DșS;{yw&PDϸ;}ZAc?t&.}R4}g<# 3~ocrVkvk,fFߺM{r儩Oy z޲#V/YAݵkV]y尼\}\ X>@3gغϬwWj)[l>ֺǵ.֬}nSR꧶U( y8i>6A+HttO! 8Pǹڗ]Ry*!H[0 f篹2ME@F)4ZmsN|u#Blh7:"vv쏀BvO}Ԟ.Q,LwJ?7N A=6ě>}p-.UCr1>JOƳYyPtݨu]r֖E+ @M#a3Z*KATov"p9X˲ObQQ4ykxk4FhT.zUab˞ (^nj@)ho-D=ח)ώ$RY~V-=KgAƯkvcq@ZH(cdH7C!=9A.qT n _CZ(,U/q J,$Tl4+71 :uꭷފKl ()jȈi1<}8̙UeH*u4"IШDM2=*\R⟪hziByVM;S7b0?/m#fՄ612Խ&w]ܡJ=~CmXM`Z! h}@igmM mTSU #XG><%t֬YA#JlXo!K$_MW~s<%Q+EuDj%}q~rP0Yey*hfo˓)L(ې7%lf rFJ3θx=6D;jp Č ~h@U׮fy5ժe^ nmYq{|)XC9L$CmRoD%Ѣ Q^"(\`YX'1*]$ yMh)+M8 y ?MIEёɌ5WZklh ,t u5P|P2hNAKkK'unxMNtSjp0:`hvS(y'Q1)7zh{ԓ/􈆼5sL 0OG=Vc9oG?OL_*hK=ַ>6 /N>Q7ϝim2e S_2 `I yW2Wl{d_LewW\^Nۈ?33򆩬E~;v-myxMGCLL4h2~V.fAvGz+!*0ԓ?Iv )%Ulə4MHy@7yDLv:Dӿo${ FC3fmaϠmAf}̓;L4*̅~"W}s#:,2xaڴixuB. 0"|ҤP|6殻eo~VD/]j<8D{l<}SLYcgXy:?JJ?.kmEba}OlidRȨtҎQ=U!?A^"xֳː^ys_d3:\QG4uI'߿/pgy;M5qO[ڽ컡?Ϯz꙽oַALov^,Ļy$GBg>{ƍw/.BTR;ttH?!MMތLi;dMAvM3gw0>e˖/NquQ%"k^CHbW_; 2> ofB]"\'[]FuwU^d03[6({L. |_?{qe;Njo7 Flh g/Zi4/&dlBRuFRNЀBB/GGawK5ܑJ} e=lGs&aNF00b[T:boΙkxٸ;k!`h6Msȧ}%l}F \K<;w.V7 pZ0~S-.W`_X>i)M0rg8(֥4f*bXt[0cxdIB0hƓR$QczOij;&,6K1lHĤFdno-q95ؗR]hl'\RVs1D_ 6to bCXW^]nV!dS'mΖLGکr5p[YdF`!)yhpL9#Y$C]Iᴨu` l_SZ.ٷ3Iɘ.l5.Mf|:iAL\yhw:Z,JԅZ[^IûQukH)#Kmw>nCZ-PԫqC,a.f= R1E9}bƎs&_tw5 "҅FLoAu"W{׷-Xf*E] O1бl)(>X@(-pޕ50~w?>SNZ[$QIu[.o}MNJ2VS[N3dz꩚P̃2[;ʌM&J)"N!c/K"D^&5qD;:>3 (F\x 0i0X*}6kAfXb :ȖM`-i^U4T*GJ2եt)oV Dwm29hh'!nETCvSYnS!+`!M8Oz־KkUax\ 2lA jZP'TN3tL|N?l(U"zE ]c{@ʌuW”2pN Z( ?VA16loق%KDP,XmViL7,)oHl"*/PQ杔>?b.Fh>呆; 4JR_҇;M@QlWsv vB炊-rԩg?\n5B$i.'4I4)JwFGӤܚ5ğ1a.vR'p.Z ))Z䑄9}BыDB>J`eL,lYJzt Ni tκhbw#/~͎ -[jП^T|2#=7+A&43"lPj)񲉍b|Y)KsdĻۄvh^qRb>)`M ̰]VU\k`#O>n1 Fd֪;Z /,O  2iiH.~,}hljǂ 7Sلj*泯bF*9#0J3reaf\4=a] 1rN|m@=JMXދ2m|@M͇ƴyuPz*g"ܕn6 #  M"AGdz{ eH[:x-5`9x{kp,%W[{6$;)2"}v> lef4*qf89Gd<_K݁M  vלK\!}J6qjlkË͉Z};lU /hF7T|R:]q"y]ؓ0CKӤVUJ"HՔk˸:u=xoW%i>u+Ɂb7=1<\9-@* 4"YSg!By2G=>%+X ICC)ZfEs"WsSfQʔϡg;@mue)|//êELsn2DuW7N%0/I w7h*rw%5%$ ~n ͧk~8R '4uӶCj.!fm̔FDa) E %LƏѓa5䈳&BdF5oII6j"t$y:]4m ¬~mxFȲrnps1Dm^b=}'}lBFn,0Bj&3~[!pbWEt Ӫ[(BHF+9^kS| `GFӶژ7iqy1Bk>}pH!hv%%T@z?q%oXh C}:S };kG|ƕb.8?^sYӂOLIK0Nմl=(;UxG>UYiV6%%A &#q$wncx^6ǷO}TI0 \tZK]z V;<.BRJH)kq(yUjʺֻ!鑧yʋ8h_$Tx]z\aky ؒlmǿfUrvەyGꞧAQ S mlFd>ځ =آ n(Nr9$7%Ք(!Ȣ84H*Z#ޜ\h68eK*2:ᮾ m׊%Ĩӫ,d-L\hC%\eZ'N~>=U1x3, &F3BD@gRȪ 2H!=㾥=u]4zM5+iP!bA1dDn4d QF/,/ҀB.PٰlbMd1R Xӑ۝d:=iC j}ApgMXMΪxVg?S2{# >8j" 42ҝ* -"6rXݍE{&=Pp]Ҥk99 I@+`HQSB$$޺f;!/a-pt=-12J& #"F`Tz$s!FL6J٨\"I=4 ym@&Aq %qjI*Pir4#:{8+[o4e肤Аw@aQ4ű~s>`stтjB` gًW6 $D5S&f2 lɵ (T>H6*O&k6T$y84w(0We^m=)rD%JTCMRTä{Ag"jkusBDzYM(;l(# :9X:9qKYL_Nٺ7 =4 Ʊݛ=/[5L,6 LhQbBY&irxEkcyERXt RpP(ݯ )93ȘBwfb.ֲCLbmIMN(3]>O;F(%ӏ4Q;WJWn*t2Q(B/6V> uA+FV,_>+\$ }W"2a((0q+<+ 8%l.]$hQ1Md ûF>X!}ހ48~ `| a8҅l)5>o1I% ث R0qZ5V_& rSZhą6EM(;Q$<(o'?D1J|c,ZQzqMa ĉ汙2߫8W=3]7}}SOe&;5~q!p&uK P#eFK򖺩~X^j(2#<7zKګK"U%ސfG nM,xDp?WMY44޴J6JLxaj+Eڔ(#Bp@ӂ֍1qjĭ˪aWwȜ%7b}8ѺܝV]Ԩ:h|ej`j&iS&W,/eU M]#F,fDd6+7;^l]üv$Mm|Źè_tm.QefGl#fcsTlWѻZg ~pDeLZ_ƴxi9^EƫxhNW*2^EƫʌNJJ*2^I֘IIENDB`n 666666666vvvvvvvvv666666>6666666666666666666666666666666666666666666666666hH6666666666666666666666666666666666666666666666666666666666666666662 0@P`p2( 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p 0@P`p8XV~$OJPJQJ^J_HmH nH sH tH J`J MNormal dCJ_HaJmH sH tH d@"d .* Heading 2ddd@&[$\$5CJ$OJPJQJ\^JaJ$`@` .* Heading 3$$@&5B*OJPJQJ\^JphOf@f  f Heading 4$$@&%56B*OJPJQJ\]^JphODA`D Default Paragraph FontRi@R 0 Table Normal4 l4a (k ( 0No List RR .*Heading 2 Char5CJ$OJPJQJ\^JaJ$`^@` .*0 Normal (Web)ddd[$\$CJOJPJQJ^JaJ*W@* .*`Strong5\T!T .*Heading 3 Char5B*OJPJQJ\^JphO6U@16 .*0 Hyperlink >*B*ph.X@A. .*@Emphasis6]RRR .*0 Balloon Text dCJOJQJ^JaJNaN .*0Balloon Text CharCJOJQJ^JaJZqZ  fHeading 4 Char%56B*OJPJQJ\]^JphO44 gQ0Header  !66 gQ0 Header CharCJaJ4 @4 gQ0Footer  !66 gQ0 Footer CharCJaJPK![Content_Types].xmlj0Eжr(΢Iw},-j4 wP-t#bΙ{UTU^hd}㨫)*1P' ^W0)T9<l#$yi};~@(Hu* Dנz/0ǰ $ X3aZ,D0j~3߶b~i>3\`?/[G\!-Rk.sԻ..a濭?PK!֧6 _rels/.relsj0 }Q%v/C/}(h"O = C?hv=Ʌ%[xp{۵_Pѣ<1H0ORBdJE4b$q_6LR7`0̞O,En7Lib/SeеPK!kytheme/theme/themeManager.xml M @}w7c(EbˮCAǠҟ7՛K Y, e.|,H,lxɴIsQ}#Ր ֵ+!,^$j=GW)E+& 8PK!Ptheme/theme/theme1.xmlYOo6w toc'vuر-MniP@I}úama[إ4:lЯGRX^6؊>$ !)O^rC$y@/yH*񄴽)޵߻UDb`}"qۋJחX^)I`nEp)liV[]1M<OP6r=zgbIguSebORD۫qu gZo~ٺlAplxpT0+[}`jzAV2Fi@qv֬5\|ʜ̭NleXdsjcs7f W+Ն7`g ȘJj|h(KD- dXiJ؇(x$( :;˹! I_TS 1?E??ZBΪmU/?~xY'y5g&΋/ɋ>GMGeD3Vq%'#q$8K)fw9:ĵ x}rxwr:\TZaG*y8IjbRc|XŻǿI u3KGnD1NIBs RuK>V.EL+M2#'fi ~V vl{u8zH *:(W☕ ~JTe\O*tHGHY}KNP*ݾ˦TѼ9/#A7qZ$*c?qUnwN%Oi4 =3ڗP 1Pm \\9Mؓ2aD];Yt\[x]}Wr|]g- eW )6-rCSj id DЇAΜIqbJ#x꺃 6k#ASh&ʌt(Q%p%m&]caSl=X\P1Mh9MVdDAaVB[݈fJíP|8 քAV^f Hn- "d>znNJ ة>b&2vKyϼD:,AGm\nziÙ.uχYC6OMf3or$5NHT[XF64T,ќM0E)`#5XY`פ;%1U٥m;R>QD DcpU'&LE/pm%]8firS4d 7y\`JnίI R3U~7+׸#m qBiDi*L69mY&iHE=(K&N!V.KeLDĕ{D vEꦚdeNƟe(MN9ߜR6&3(a/DUz<{ˊYȳV)9Z[4^n5!J?Q3eBoCM m<.vpIYfZY_p[=al-Y}Nc͙ŋ4vfavl'SA8|*u{-ߟ0%M07%<ҍPK! ѐ'theme/theme/_rels/themeManager.xml.relsM 0wooӺ&݈Э5 6?$Q ,.aic21h:qm@RN;d`o7gK(M&$R(.1r'JЊT8V"AȻHu}|$b{P8g/]QAsم(#L[PK-![Content_Types].xmlPK-!֧6 +_rels/.relsPK-!kytheme/theme/themeManager.xmlPK-!Ptheme/theme/theme1.xmlPK-! ѐ' theme/theme/_rels/themeManager.xml.relsPK]   +++. Mo=17DG/IRNW]h`bfm~G3 N[]^`acegmuwxz|}~z:.^5= >]>@AE2FgFFFHGGHRVdAq3MN\_bdfhijklnopqrstvy{%,*@G[7u| 5MTX " ) 1 Q X Y ~  d p v h z DK)ELnR[ }  lw"ɧ!m%Ъ'rXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX $'.!d0# AA@ D (    /@8  "0? G /  s"*,%/@8h ,/ +  "/c  h@ABC/ EF/  @`#" )   h@AB C EF9r 9@`#" ,x) +   h@AB7C EF 7 7@`#" x),/ +   b@ABCa EFf a @`#" /F   b@ABCEF @`#" G    b@ABCEF~ 0@`#" G   h@ABC EF3   @`#" !-,  h@AB{CEF3{E {@`#" !  h@ABCh EF3k h @`#" /c  S s"*(g  z  S #" ^+,%.   S c"$(Z% B S  ?vF.*2b#ContentsIntroduction_to_neural_networksWhat_is_a_Neural_NetworkHistorical_backgroundWhy_use_neural_networks(Neural_networks_versus_conventional_comp(Human_and_Artificial_Neurones_-_investigHow_the_Human_Brain_Learns(_From_Human_Neurones_to_Artificial_NeuroAn_engineering_approachA_simple_neuron _Firing_rules Pattern_Recognition_-_an_exampleA_more_complicated_neuronArchitecture_of_neural_networksFeed-forward_networksFeedback_networksNetwork_layersThe_Learning_ProcessTransfer_Function(An_Example_to_illustrate_the_above_teachThe_Back-Propagation_Algorithm(Appendix_B_-_The_back-propagation_AlgoriApplications_of_neural_networksNeural_Networks_in_Practice_Neural_networks_in_medicine(Modelling_and_Diagnosing_the_CardiovascuElectronic_nosesInstant_PhysicianNeural_Networks_in_business MarketingCredit_Evaluation Conclusion References_PictureBulletst !))g-j//1:LBFFHOK%Udj8qxFj7ċ)m   !"$!)*-//1':eBF GH]K9UdjVqxeR:~! $$--k.s...U9c90B6BKKPP QQQQQQQQR(RRRRR0S;SSTTTTUGWSWUWWWYY]]]]]][^]^^^`^a^c^r`t`bb/g7gyyyyyyzz||;|=|[|^|~||||||||}}}}~~3~6~/9jtv@FSWǩީ_ervx !""$$%%'(*+-.MNNRN/U/e/i/%414I5L577<<<<GCRCDDBQJQQQYYSZZkbbbbop5|:|tw !""$$%%'(*+-.MN333333333333333333333F8xG3-N-P-d-3035]689=?DDFFH'H-HKH2L@LnPrP }G}"03Sp#;g !""$$%%'(*+-.MN8.:Pܚ@7G4]MbyuziyWzBh ^`.^`.pp^p`.@ @ ^@ `.^`.^`.^`.^`.PP^P`.^`.^`.pp^p`.@ @ ^@ `.^`.^`.^`.^`.PP^P`.^`.^`.^`.  ^ `.\ \ ^\ `.,,^,`.^`.^`.^`.^`.^`.^`.  ^ `.\ \ ^\ `.,,^,`.^`.^`.^`.^`.^`.pp^p`.@ @ ^@ `.^`.^`.^`.^`.PP^P`.4]Mbuzi8.:@7GWz=?c9`ddT0ddc&jIdd@ rgI/PddQ>/6"ddTSNla/PX\x[:>/6"dd1\/6"dd'>/6"^ g"x[:ddT}",s'/6"dd.;/o1NS24ahU7+7I;?/6"ddYy>Mndd=3Nz.QQ>/6"ddi\0 Ddd$QN]!fddp^h^@ dds_^)`Xbmqc6LfT}"dd!f>/6"ddxi'ddX$k=3Nddn>/6"ddrpT}"ddr_uDS+dd;+y?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[]^_`abcdefghijklmnopqrstuvwxyz{|}~Root Entry F`z_lData /1Table\NWordDocument SummaryInformation(DocumentSummaryInformation8`$MsoDataStore k0tl4W23EZUUZQ==2pSk0tlItem  PropertiesUCompObj y   F'Microsoft Office Word 97-2003 Document MSWordDocWord.Document.89q