ࡱ> ]_Z[\s` <bjbj M%    ~k~k~k8k<kT Roho"ooorrrȉʉʉʉʉʉʉ$h'fsq^rfsfs  ooÊLvLvLvfs^ ooȉLvfsȉLvLv6LvoFo tZ~ksLv|vLي0 Lvt0LvLv0rLSr6Lvr,rrrru^rrr fsfsfsfsf~k~k       Module 5 Forecasting (These notes are new and probably contain errors. Dont hesitate bringing them to my attention. If a formula or values in a table seem incorrect to you, let me know and Ill look into it.) Introduction A forecast is a prediction of future events used for planning and decision making purposes. Managers may need forecasts to anticipate changes in prices or costs, prepare for new laws or regulations, track competitors, or analyze resources. Keep in mind that, while forecasting methods provide useful information for planning purposes, they are rarely perfect and must be used in conjunction with other information. Forecasting methods may be based on mathematical models using historical data available or qualitative methods drawing on managerial experience. In this module we will explore several forecasting methods commonly used today. Components of Demand The forecasting of customer demands is at the root of most business decisions. It is an inexact science as the demand for goods and services varies greatly and can be affected by unforeseen circumstances. Forecasting demand requires uncovering the underlying patterns from available information. This section discusses the basic components of demand. The repeated observation of demand for a product or service in their order of occurrence for a pattern is known as a time series. The five basic components of most demand time series are: 1. Average, or the sum of the demand observations for each period divided by the number of periods. 2. Trend, or systematic increase or decrease in the average of the series over time. 3. Seasonal influence, or the predictable increase or decrease in demand depending on the time of day, week, month, or season. 4. Cyclical movements, or the less predictable increases or decreases in demand over longer periods of time (years or decades). 5. Random errors Cyclical movements arise for either the natural business cycle or the product (or service) life cycle which reflects the stages of demand from development to decline. Business cycle movement is difficult to predict because it is influenced by so many national and international events. Predicting the rate of demand buildup or decline in a products life cycle is also difficult. The ability to make long range forecasts depends on accurate estimates of cyclical movement; we are going to concentrate our efforts on short and medium range forecasts. The four components of demand - average, trend, seasonal influence, and cyclical movements - combine to define the underlying time pattern of demand for a product or service. The fifth component, random error, results from chance variations and thus cannot be predicted. Random error is the component that makes every forecast incorrect. There are three basics methods of making forecasts judgment methods, causal methods (linear regression), and time series methods. Judgment Methods These are qualitative and typically used when historical data are lacking such as when a new product is launched or technology is expected to change significantly. Four commonly use judgment methods are listed below. Sales Force Estimates Often the best information about demand comes from the sales force. They are likely to know which products customers are buying and in what quantities. The information is usually readily available by districts or regions and the data can be easily aggregated. However, sales force estimates may be biased. Individuals may not be able to differentiate between customer wants and needs. If the firm uses sales as performance measures, underestimating demand may occur. Executive Opinion High level execs are paid multi-millions to anticipate customer demands. Market Research This is a more systematic method involving questionnaires, surveys, sampling, and information analysis. It is best left to the experts in the marketing department. The IT department may be called upon to assist in gathering the data (developing an online format for a survey) or in the presentation of results. Delphi Method a process of gaining consensus from a group of experts while maintaining their anonymity. A coordinator sends data and questions to the experts, their comments are shared among the group, discussed, and, hopefully, some consensus is reached. The process typically takes a long time. Causal Methods (Linear Regression) Linear regression can be a course unto itself. Only a barebones discussion is presented here and should be a review from some previous college course. This causal method is usually very good for predicting turning points in demand and preparing long range forecasts. In simple linear regression one variable, called the dependent variable, is related to one or more independent variables by a linear equation. The relationship is easily graphed using a scatterplot in Excel. A common use of linear regression is using SAT scores to predict college GPA.  Let the independent variable be SAT scores and the dependent variable be first year college GPA. The eleven points in the scatterplot represent historical data (SAT, GPA) for eleven students. The next step is to find the line of best fit or regression line. The math is fairly complex but Excel computes it automatically and you only need to interpret the results. Lets use Excel to conduct a simple linear regression on a sales-advertising problem. A manager must schedule production to meet anticipated demands for a product. The following are sales and advertising data for the past five months. The marketing manager says that next month she will spend $1750 on advertising (the independent variable). We can use the linear regression tools in Excel to forecast sales for month 6. By convention, the independent variable is graphed on the x-axis so when setting this up in Excel, Advertising should be in the first column. Month Sales (000 units) Advertising (000$) 1 264 2.5 2 116 1.3 3 165 1.4 4 101 1.0 5 209 2.0 Construct a basic scatter plot then select the chart and select Chart-Add Trendline from the menu options. In the Add Trendline dialog box select Type Linear and select Options- Display equation on chart and Display R-squared value on chart. Your results should look something like this.  EMBED Excel.Chart.8 \s  Now what? Use the regression equation Y = 109.23X 8.135 to forecast Sales (Y) from Advertising (X) Forecasted sales Y = 109.23 * 1.750 8.135 = 183,018 units. What about the R2. (Note some texts use r2.) The mathematical name for R2 is coefficient of determination. It indicates the proportion of the variation of the dependent variable that is explained by the regression equation. For your purposes, it ranges from 0.00 to 1.00 and values close to 1.00 are desirable. You are probably more familiar with the term, correlation. Correlation measures the strength of the relationship between the independent and dependent variable and can be calculated by taking the square root of the coefficient of determination, R2. In this example, the correlation coefficient, R (or r) = .98. Time Series Methods 1. Simple Moving Averages This method is used to estimate the average of a demand time series by attempting to remove the effects of random fluctuations. The calculations are straightforward and involve finding the average demand for the n most recent time periods and using that average as the forecast for the next time period. Moving averages are common in forecasting share prices in stock market analysis; you will find these averages shown on most brokerage sites. An example applied to weekly patient arrivals (demand) at a clinic is included on the spreadsheet on the next page. A formula for the 6-week moving average of patient arrivals might look like this: Ft = (At-1 + At-2 + At-3+ At-4 + At-5 + At-6)/6, where Ft is the forecasted patient arrivals for the next time period, At-1 is the historical patient arrival data for the most recent week, At-2 is the historical patient arrival data for the second most recent week, etc. 2. Weighted Moving Averages In a simple moving average each demand has the same weight. In a weighted moving average forecasting, each demand can be assigned a weight relative to the other demands making up the average. The sum of the weights must be 1.00. The advantage is that it allows you to emphasize recent data over earlier data. The weighted moving averages in the following spreadsheet were calculated using a weight of 0.5 for the most recent, 0.3 for the second most recent, and 0.2 for the third most recent. The patient arrivals for the next time period, Ft, are calculated as follows: Ft = .5 *At-1 + .3*At-2 + .2*At-3  WeekHistorical Demand3-week MA forecast6-week MA forecastWeighted MA forecastExponential smoothing a = .1Exponential smoothing a = .4140023803411390.0390.04415397399.5392.1398.45395402406.8394.4405.06375407404.2394.5401.07410395396389.0392.5390.68397393398396.5394.3398.49407394401396.5394.5397.810422405400404.6395.8401.511430409401412.5398.4409.712392420407423.0401.6417.813396415410409.4400.6407.514415406407401.6400.1402.915387401410404.7401.6407.716401399407397.2400.2399.417389401404399.6400.2400.118407392397392.2399.1395.619398399399400.4399.9400.220427398400398.9399.7399.321412411402414.3402.4410.422378412406413.7403.4411.023390406402398.0400.9397.824408393402390.8399.8394.725429392402396.6400.6400.026408409407414.9403.4411.627393415404414.3403.9410.228409410401404.7402.8403.329 403406404.0403.4405.6 3. Exponential Smoothing Exponential smoothing is a more sophisticated weighted average method that calculates the average of a time series by giving the recent demands more weight than earlier demand. It is the most frequently used formal forecasting method because of its simplicity, small amount of data needed to support it, low cost, and the ability to be extended for more complex situations. Unlike the weighted average which needed n periods of past data and n weights, exponential smoothing needs only three values to forecast a value for the next time period the forecast for the current period, the demand for the current period, and a smoothing parameter, alpha (), where 0 d"  d" 1. Here is the formula: Forecast for the next time period, Ft+1 = *(demand for current time period) + (1  )* (forecast calculated for current time period) =  *Dt + (1  )*Ft Here is how the values were arrived at for the exponential smoothing column in the table. Assume  = 0.1 and let s find forecast a demand for week 4. The exponential smoothing method requires estimating an initial value for the average. Suppose we use demand data for weeks 1 and 2 and average them, obtaining 390, and use this as the initial value for Ft which will be F3 in this first calculation. Then Ft+1 =  *Dt + (1  )*Ft = F4 = .1 *D3 + (1  .1)*F3 = 0.1*411 + .9*390 = 392.1 Once the first one is completed the rest is automatic. F5 = .1 *D4 + .9*F4= .1*415 + .9*392.1 = 394.4  4. Trend Adjusted Exponential Smoothing Because of its simplicity, exponential smoothing is at a disadvantage when the underlying average is changing such as a time series with a trend. However, the approach can be modified to account for trends. A trend is a time series is a systematic increase or decrease in the average over time. To improve the accuracy of exponential smoothing we include an estimate of the trend into the calculations. The trend-adjusted exponential smoothing method uses the following formula, with two parameters, alpha (a) for the average and beta (b) for the trend. Average for current period = a*(demand for current period) + (1 - a)*(Average + Trend for previous period) or At = a*Dt + (1  a)*(At-1 + Tt-1) Trend for current period = b*(average for current period  average from last period) + (1  b)*trend estimate last period = Tt = b*(At  At-1) + (1  b)*Tt-1 Forecast for next period, Ft+1 = At + Tt Again, getting started requires estimating initial values. Analyzing historical data (not provided) from the previous 4 weeks, the average demand was 28 with an average increase of 3 per week. Therefore we can let D0 = 28, F0 = 28 and T0 = 3. There are lots of ways to make these initial estimates but anything reasonable is ok because they have little impact on future results as time passes. The results for an example are in the table below. We will answer the question on how to choose alpha and beta values shortly. WeekActual DemandSmoothed AverageTrend EstimateForecast alpha=.2 beta=.20*28283 12730.202.8431.0024435.233.2833.0433738.213.2238.5143540.142.9641.4355345.083.3643.1063846.352.9448.4475750.833.2549.2986155.463.5254.0893954.992.7258.99105557.172.6257.72115458.632.3859.79125259.212.0261.02136060.991.9761.24146062.371.8662.96157566.382.2964.231668.67* not actual data but estimated from historical data 5. Seasonally Adjusted Forecasting The final technique we need is adjusting for seasonal influences. You can see the high demand in the third quarters and low demand in first quarters; it is especially evident in the graph. This might be sales of leaf rakes for example high in the fall and low in the spring. quarteryear 1year 2year 3year 4145701001002335370585725352059083011604100170285215Total1000120018002200Ave/qtr250300450550  Here are the steps to forecast the demand for each quarter of year 5. 1. Determine the yearly demand for year 5. Over the last 4 years the average increase has been 400 units ((2200-1000)/3 = 400), therefore it is reasonable to expect a demand of 2600 in year 5 or 650 in each quarter of year 5 if seasonal influences are not taken into account. 2. Determine the seasonal index associated with each quarter by dividing the quarterly demand by the average demand per quarter for that year. Then, calculate the average index for each quarter quarteryear 1year 2year 3year 4average1 0.18 0.230.220.180.2021.341.231.301.321.3032.081.971.842.112.0040.400.570.630.390.50 3. Finally, multiply the expected demand per quarter (without seasonal influences) by the seasonal index for each quarter to obtain a forecast for each quarter of year 5. quartercalculationsforecast, year 51650*.201302650*1.308453650*2.0013004650*.50325 Choosing a Time Series Method Now we have many choices so how do we decide which method and parameters to use? Forecast Errors Forecasts almost always contain errors and they can be classified as either bias errors or random errors. Bias errors are the result of consistent mistakes in which the forecast tends to be consistently high or low. The other type of error, random error, results from unpredictable factors that cause the forecast to deviate from the actual demand. It is desirable to minimize both types of errors by selecting appropriate forecasting models, but eliminating all forms of errors is impossible. Before errors can be minimized, they must be quantified. A forecast error is simply the difference between the forecast and actual demand for a given time period and may be expressed mathematically by Et = Dt - Ft There are several ways to measure the forecasting errors but we are only going to look at the Cumulative Sum of Forecast Errors (CFE) and Mean Absolute Deviation (MAD). CFE is simply the sum of the forecast errors, Et. MAD =  |Et |/n (Note: the Excel function =AVEDEV finds the MAD) The benefits of using these two measures is that the CFE is effective in measuring bias, the higher the CFE, the more bias the forecast is, while the MAD is effective in measuring the overall accuracy of the forecasting method. Lets apply these measures to an earlier example. As can be seen in the below table choosing the best forecasting method is not always straight forward. It is desirable to use the one with the lowest CFE and MAD. In this example no one method meets that criterion. The 6 week MA and ES (a=.1) have the lowest MAD but the highest CFE, while for the weighted MA and 3 week MA just the opposite is true. A reasonable compromise might be exponential smoothing with a=.4 which has the middle value for both measures. Two analysts might choose different methods depending on the situation. It is also possible to try other values for weights and parameters and observe the behavior of the CFE and MAD.  If your answers differ from these slightly it is probably because different rounding techniques were used. Assignment Complete any 3 for a B-, any 4 for a B, any 5 for a B+, any 6 for an A- and all 7 for an A. 1. Go to http://www.ship.edu/~mtmars/StatU/StatU.html Randomly select at least 30 students and interview them. Use the data gathered to construct a scatterplot of study hours (independent variable) against GPA. Include R2 and the regression equation on the chart. If an incoming student says he studies 15 hours a week, what GPA would you predict for him. Comment briefly on the strength of the relationship (correlation). 2. Get about 6 month historical data on a stoc i j  m~)L !!!!!!A!z!!!!!!!!###$7$8$R$(%)%&&&&&&&&&&&'''''U'Y''''û h .H* h .6 h .>* h .H*hjh .Uj1? h .CJUV h .\aJjh .Uh .CJaJ h .5h . h .aJh .5CJaJ@ i j  YXijm~Y^$a$I<n<<YZNO()LMZ[}~  EI px@@hh7$8$H$^h`h$a$ !!#$7$8$R$S$&&&'' ( (****** $ px@@a$ px@@^ px@@ px@@hh^h`h' (\*^************++++,,,,,,2,3,P,Q,n,o,,,,,,,,,--3-4-U-V-w-x--------.!.".C.D.e.f.........//1/2/S/T/u/v///// h .CJh .CJOJQJ^Jh .CJOJQJ^Jj h .U h .H*h . h .>*P*****+D+j+++, $$Ifa$ px@@ ,,, , , ,,,MDDDDD>$If $$Ifa$kd$$IflD֞pUy ;7?G#'$:644 la,,,,,,,,G>>>>> $$Ifa$kd$$Ifl֞pUy ;7?G#''$''':''644 la$If,,,,,#,$,%,G>>>> $$Ifa$kd$$Ifl֞pUy ;7?G#''$''':''644 la$If%,&,,,2,3,5,9,=,Dkdh$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$=,>,D,J,P,Q,S,W,Dkd6$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$W,[,\,b,h,n,o,q,Dkd $$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$q,u,y,z,,,,,Dkd $$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$,,,,,,,, $$Ifa$,,,,,,,,MDDDDDD $$Ifa$kd!$$Ifl֞pUy ;7?G#''$''':''644 la,,,,,,,,Dkdn"$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$,,,,,,,,Dkd<#$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$,- ------Dkd $$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$-!-'---3-4-7-;-Dkd$$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$;-?-C-I-O-U-V-Y-Dkd%$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$Y-]-a-e-k-q-w-x-Dkdt&$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$x-{------- $$Ifa$--------MDDDDDD $$Ifa$kdB'$$Ifl֞pUy ;7?G#''$''':''644 la--------Dkd($$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$--------Dkd($$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$----... .Dkd)$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$ ....!.".%.).Dkdz*$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$).-.1.7.=.C.D.G.DkdH+$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$G.K.O.S.Y._.e.f.Dkd,$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$f.i.m.q.u.{... $$Ifa$........MDDDDDD $$Ifa$kd,$$Ifl֞pUy ;7?G#''$''':''644 la........Dkd-$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$........Dkd.$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$........DkdN/$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$../ /////Dkd0$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$///%/+/1/2/5/Dkd0$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$5/9/=/A/G/M/S/T/Dkd1$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$T/W/[/_/c/i/o/u/ $$Ifa$u/v/y/{/////MDDDDDD $$Ifa$kd2$$Ifl֞pUy ;7?G#''$''':''644 la///////2DBBBBBkdT3$$Ifl֞pUy ;7?G#''$''':''644 la $$Ifa$/S1T1n1o1 3&333 446666D7J7Z7\7t7x7~7777770818:8;8C8D8d8e8f8h88 ;";Z;\;;;<<^<`<f<h<l<n<~<<<<<<<<\=^============= >>>>>>>>>>? ?)@h .OJQJ h .>*j"4h .U h .H* h .6h .V2444c8d8f8g8h888~;;<<==>??(@)@.@<@N@]@w@ $$Ifa$)@w@x@@@@@@@@@@@@@AA&A'A=A>ATAUAlAmAAAAAAAAAAAAAAA-B.B/BRBkCCCCCCCCCCCCCDDDDDD7FcFdFlFmFFFFFFFFFFFFFFjh .CJOJQJU^JhS\j0Uh .U h .>*h . h .CJh .CJOJQJ^JOw@x@{@~@@@@sjjjjj $$Ifa$kdC$$Iflr 4<644 laA@@@@@@@sjjjjj $$Ifa$kdC$$Iflr 4<644 laA@@@@@@@sjjjjj $$Ifa$kdD$$Iflr '''4'<644 laA@@@@@@@sjjjjj $$Ifa$kd@E$$Iflr '''4'<644 laA@@@@@@@sjjjjj $$Ifa$kdE$$Iflr '''4'<644 laA@@@@@@@sjjjjj $$Ifa$kdF$$Iflr '''4'<644 laA@@@@A AAsjjjjj $$Ifa$kdzG$$Iflr '''4'<644 laAAAAAA A&Asjjjjj $$Ifa$kd8H$$Iflr '''4'<644 laA&A'A)A,A2A7A=Asjjjjj $$Ifa$kdH$$Iflr '''4'<644 laA=A>A@ACAIANATAsjjjjj $$Ifa$kdI$$Iflr '''4'<644 laATAUAXA[AaAfAlAsjjjjj $$Ifa$kdrJ$$Iflr '''4'<644 laAlAmApAsAyA~AAsjjjjj $$Ifa$kd0K$$Iflr '''4'<644 laAAAAAAAAsjjjjj $$Ifa$kdK$$Iflr '''4'<644 laAAAAAAAAsjjjjj $$Ifa$kdL$$Iflr '''4'<644 laAAAAAAAAsjjjjj $$Ifa$kdjM$$Iflr '''4'<644 laAAAAAAAAsjjjjj $$Ifa$kd(N$$Iflr '''4'<644 laAAAAAAAAsjjjjj $$Ifa$kdN$$Iflr '''4'<644 laAAAAAAAAsjjjjj $$Ifa$kdO$$Iflr '''4'<644 laAAA-B.B/Bsj*(?kd Q$$Ifl'644 laA $$Ifa$kdbP$$Iflr '''4''<644 laA/BRBSBjCkCsCzCCCCCCChkdQ$$Iflr ]-'''''644 la $$Ifa$ CCCCCCCCCCjkd R$$Iflr ]-'''''644 la $$Ifa$ CCCCCCCsjjjjj $$Ifa$kdR$$Iflr ]-'''''644 laCCCCCCCsjjjjj $$Ifa$kdS$$Iflr ]-'''''644 laCCCCCCCsjjjjj $$Ifa$kdS$$Iflr ]-'''''644 laCCCDD DDsjjjjj $$Ifa$kd$T$$Iflr ]-'''''644 laDDDDDYDZDpEqE6F7F?FFFsqqqqqqqqqhh $$Ifa$kdT$$Iflr ]-'''''644 la FFMFTF[FcFdFfFgFlFWQF $$Ifa$K$$Ifkd[$$IflKֈb n2'b'H'b'b'b'b2644 lal $$Ifa$lFmFnFsFxF}FF $$Ifa$$If?kd*\$IfK$L$KP'P634aFFFFFFFF`WWWWWW $$Ifa$kd\$$IflKֈb n2'b'H'b'b'b'b2644 lalFFFFFFFF`WWWWWW $$Ifa$kdB]$$IflKֈb n2'b'H'b'b'b'b2644 lalFFFFFFFF`WWWWWW $$Ifa$kd]$$IflKֈb n2'b'H'b'b'b'b2644 lalFFFFFFFF`WWWWWQ$If $$Ifa$kd^$$IflKֈb n2'b'H'b'b'b'b2644 lalFFFFFFFF`WWWWWQ$If $$Ifa$kd@_$$IflKֈb n2'b'H'b'b'b'b2644 lalFFFFFFGGGGGGGGG H H H*H}HHKKLKPKQKUKWKdLfLLLLLPP^Q`QRRRRRRSS S_T`T[U   !'(./56=>˽U h .H*jhdfUmHnHuhdfh .OJQJh h .H* h .>* h .5h .h .CJOJQJ^J h .CJHFFFFFFFF`ZZZZZZ$Ifkd_$$IflKֈb n2'b'H'b'b'b'b2644 lalFFFFFFFFFGG`^^^^^^^^^kd`$$IflKֈb n2'b'H'b'b'b'b2644 lal GGGGGGGGekd>a$$Ifl;F 6'G''e6    44 la= $$Ifa$GGGGG $$Ifa$ekda$$Ifl;F 6'G''e6    44 la=GGGGG $$Ifa$ekdFb$$Ifl;F 6'G''e6    44 la=GGGH H $$Ifa$ekdb$$Ifl;F 6'G''e6    44 la= H H H H*H+H|H}HHHJJWKXKRRR$a$ekdNc$$Ifl;F 6'G''e6    44 la=RRRRRRRRRRRR S!S}S~S,U-Ujk$$If^`a$ ^`k or mutual fund of your choice. Use an Excel spreadsheet to chart the closing prices for the six month period. On each scatterplot include two simple moving averages, say 15 and 30 day, for example, and one weighted average, with at least three weights of your choosing. (If you dont have a favorite site to get financial data try http://chart.yahoo.com/d) 3. Get about 6 month historical data on a stock or mutual fund of your choice, but different from the one selected in 2. above. Use an Excel spreadsheet to chart the closing prices for the six month period. On the chart include two exponential smoothing forecasts with different alphas of your choosing. 4. The following data are available for the monthly demand of a product: PeriodDemand121522503275428553506285730583009300103801135012325 Using Trend Adjusted Exponential Smoothing with a = b = 0.2, determine the forecast for the next periods (i.e. 13). Suppose that the actual demand for period 13 turned out to be 395. What would be your forecast for period 14 at the end of period 13? What will be your forecast for period 15? 5. Use seasonally adjusted forecasting techniques on the data below to forecast the demand for each month in year 4. Include a chart of the historical and forecasted demand. MonthYear 1Year 2Year 3January232849February213252March243445April263843May223144June273240July254046August273848September244254October234450November294548December314756 6. Duplicate the values and interaction on the spreadsheet found at: http://www.ship.edu/~mtmars/MIS_530/forecasting/int_forecasting.xls Use it to determine the best combination of alpha and beta. 7. For the data below calculate the CFE and MAD for at least 7 different forecasting schemes. Include at least two each of simple moving average, weighted moving average, and exponential smoothing with different time periods, weights, and/or parameters as appropriate. Include a table of the calculated CFE and MAD measures and a brief discussion of which forecasting scheme is most appropriate and why. HRkdNd$$Ifl0! ''644 la$$If^`a$Rkdc$$Ifl0! ''644 la  HRkdFe$$Ifl0! ''644 la$$If^`a$Rkdd$$Ifl0! ''644 la HRkd>f$$Ifl0! ''644 la$$If^`a$Rkde$$Ifl0! ''644 la !#'(*.HRkd6g$$Ifl0! ''644 la$$If^`a$Rkdf$$Ifl0! ''644 la./1569=HRkd.h$$Ifl0! ''644 la$$If^`a$Rkdg$$Ifl0! ''644 la=>AEFIMHRkd&i$$Ifl0! ''644 la$$If^`a$Rkdh$$Ifl0! ''644 la>EFMN              ' ( 6 7 G H [ \ m n         <<<<<<<<$<%<+<,<2<3<9<:<@<A<G<H<I<J<L<M<O<P<R<S<U<V<\<]<^<`<a<g<¼¼¼ h .0Jjh .0JUhCjhCUUh[hS\h .OJQJh . h .CJh .CJOJQJ^JJMNO          $$ W&#$/If^`a$ ^`Rkdi$$Ifl0! ''644 la      }cccc$$ W&#$/If^`a$kdj$$Ifl\? '''' 6 W644 la      }cccc$$ W&#$/If^`a$kdj$$Ifl\? '''' 6 W644 la      }cccc$$ W&#$/If^`a$kdRk$$Ifl\? '''' 6 W644 la      }cccc$$ W&#$/If^`a$kdk$$Ifl\? '''' 6 W644 la      }cccc$$ W&#$/If^`a$kdl$$Ifl\? '''' 6 W644 la   ! $ ' }cccc$$ W&#$/If^`a$kd m$$Ifl\? '''' 6 W644 la' ( - 0 3 6 }cccc$$ W&#$/If^`a$kdm$$Ifl\? '''' 6 W644 la6 7 > A D G }cccc$$ W&#$/If^`a$kdTn$$Ifl\? '''' 6 W644 laG H R U X [ }cccc$$ W&#$/If^`a$kdn$$Ifl\? '''' 6 W644 la[ \ d g j m }cccc$$ W&#$/If^`a$kdo$$Ifl\? '''' 6 W644 lam n w z }  }cccc$$ W&#$/If^`a$kd"p$$Ifl\? '''' 6 W644 la      }cccc$$ W&#$/If^`a$kdp$$Ifl\? '''' 6 W644 la              }ssssssssssss ^`kdVq$$Ifl\? '''' 6 W644 la   + i j << <<<<<Rkdq$$Ifl0u 5 ''644 la $$If^`a$^ ^` PeriodDemand1100280311041155105611071258120     PAGE  PAGE 1 45/250 = .18 <<<<< <$<HRkdr$$Ifl0u 5 ''644 la $$If^`a$Rkdlr$$Ifl0u 5 ''644 la $<%<'<+<,<.<2<HRkds$$Ifl0u 5 ''644 la $$If^`a$Rkdds$$Ifl0u 5 ''644 la 2<3<5<9<:<<<@<HRkdt$$Ifl0u 5 ''644 la $$If^`a$Rkd\t$$Ifl0u 5 ''644 la @<A<C<G<H<I<K<H>< h^h`Rkdu$$Ifl0u 5 ''644 la $$If^`a$RkdTu$$Ifl0u 5 ''644 la K<L<N<O<Q<R<T<U<^<_<`<k<l<m<n<o<p<~<<<<< h^h`h]h&`#$g<h<i<j<k<m<n<p<|<<h .CJOJQJ^JhCh . h .0Jhdf0JmHnHujh .0JU (/ =!"#$% nsXp] PNG  IHDRBM <sRGBIDATx^ `g9*DmVV۪mm^ZZZAm`Q 땨jPD3ܛ$do-03;fvv滌qv>oޛ7lWR}r{S>'EQ0>=zss_Ñ_>>      E樃.9Q&CKK># Qo~绨>7;JX87ܻ r#yw/n)@@@@@!@Qu#~wvT*U [EpKG=(.-Ig- uKb|y˗c^kAW޾x0s}ڑJѮ>    V4h 2sB<78O+WZ~^yL0')d;M>DBwOLZmMq; k}Z'Ƈ(![<HLcj{4 tY0 ̾ 2G^m FRW(JL4fEؤ`| )A{G>M߻3^c]9qC!o _QH^wMަgvGZh+)ˢ3S+ O1//WjjZ. "rJ]a9 KѶ-T(m|o%*Iyo mQ@|c!C}y}M@& vvʪC+*QHGCS4jm 45mQxsN|Kھ1;cGMo{)@[Wv%6#PW_r4~ X<d@3> YF :Ы?"'7׷%뢠´E;BX{Gt_޶jn! !' 4z̫{}/77ܴb0R, |0 (tuX=ܶ4Y}%w>/j:(Q0HG=+_ ,C/./:DeY|     ^g]ȥhs s#`4 Kޢ4(Quv =Vo{kK71NrBr/X/:ΗG#`͑hG&ZRFihF"S7,f涶/wx6av=ɥ9Zr+t)WDE_ JouDi     @ P*ݻOڂE[i^t_k&%DidlOs4/mGr|yHȼ_o{=wkn}^"ˉ>YE{)*8%mm 6G=F/'Ǘ[0_^6' ~V:tMϏ9re㯜ҏ;iԌf!i;*_=㡻+VXv}j{4OIܳX)k0_p`%Z$I̩]c5:XWüDpd     tA{ȫntJ$=E`A0/=Ɵe hi yvwmIV7q+m="}1[|=Efȏ~uuf;aɷ+m nˑrR4ص~>qQîE} _"Yc[CCaLVND#`Qyb0Z^"9좿{U&T_TmC~OJ_cG c)Yb7kA:$jX%R27C7l%h yCytK~+"<`4/o`Gs(?P7onoE׷]vO(tVih#Bs)~4vp#/s_JЁkX:'$#>MA7 ?zrc::c}nZIçD-H#ðXn/.ϻ}.t7q3mI,8:(_[NLd     8`g|zƖmChXW𩣞qy5^/]߼ϻtN_&>BhHlzx}}MZpKMguk3GʧW=*"rLEO{3l/^=H    B(_k 5%ɉJeSriEr}fmQc,0xaBc$_ IFX^.z&mؿ-ɉ| 鋆|bu?Ѫ\)bp~5{=xn}F? 7wF%z,O͕7_|!<8A}1G;\>`dJ^%48yJ| _.hҳ|ÂmQ_NvF+BV~"Usd qzl    _%^f!C9Wng&"ջPR8/*f/h_W,_|G;=6on?&y!" %=="!X^;YNG">"hlv"}zȥy {S|~aEEnj?_82+[tt(Z{ΡiP`?~TٗM}Ӝ+1;S'Ĕ}dq%rDL-2ԊiB!CfZm39shObc?(WNZ>H' z`beB^,y* \:y0GFq0Em\C裉!,viqЁwBd&U~U)HFb%JLǙ_,L*u5@@@@'@4[t^^k[zpkno;HX| "=zE_3cJB_ckxȫ{(*.c ǖ3"!s^`| znOK WS^c'tvvvh=G&hh/L[.G_<"fR^/'Q#'D_WuWE Kge,Ď&g䔖{JhVL9HJ;jKy(OI}zj C Jб"4H#c=|w!žR|[wk j>:R^LB--59Wx|gE YN_(HC^6ͳ v54--` G{eG}Pr) G SҘb:;DjRtCs4iD:'bk ~ @@doeJ X&<< D}ةF%|`8ݱ:Y91KB|ߋ멞[\]ӳ/?%YAEQzȫ-%Z&;a԰' H*GVg;߶Lγ#W|3-زlEd@@@2GXأdH!_})S׳r0BiIAy'G_<pYEtѹ? %'7-LT>fRqnAQNoI/.|!`Ljj']U4X, J=}rimZP^`#\y]0A4` xƝMW~ܼ[W ӛ(͗K3hCZӢ+죗i/8`G'&!`-sޝR~|Fh1z҆Fhplyzct u:{lD?7׻)Gچlu ZiC> b;س`4&Й XVVt"LG   &9wƑѩ|e{K^!=%/B/5rhp(¤\|F諭w_Ks&dP%aϧ6mے#YE"ƹ<-hZQփF:aGh3_WIe_NPJ^0 -(d˦0_Yux@y7GۗʒT%]S*@)[#]ry͝r''|%"kZv=(K=s{|=wJkXFsze= gr0棷X"kb2[>Eb,Fp NgL.hč__߃}(#2O/ iD2!<)@o^ qh3Wм^&iͥW0ˡX<9R}y0)lXB.ZBE! D @EPM4|*zUM'b(WG3ƷBFXUYrEQ8䗌x{p@p0r@{HEH_jo:k01ő|Zx3Oe?gV~W)Jʑ81G{3+.oJzKۃSE?)a ڤ|YR%ʨs+Ӛjz^%C?2g53 = ^==cyzk)z<3}Nl}kҁ       zR%@yOܧ?n1[;uKuL&/N{piD:'(Fl󷉒'eeS g]‰F6 &ʕ(WF11q2k DSg=+WuRtJs5@@p\Ck^"P+hhhh :^,kO[CN:ޥ0{";NwAuq`~AkGE^fjN}0;ΞL/%ݚvA {K#`KXzyf\,t;gQA7z;+5,ѣ)/0ÏeCv>clL͒#ocXÈiM5[YUxfgp .R1f{@&S-+0glK^;2>-ye \KOl\s.>,d7# O: M,  BP/%YX.Q~` { }Zomo~Nuq{w;ꥮ^]=k(pEzU(# e^MJ~LO'of?)[f5QJvqm>I?Kb[Qc^\8/2pP+}!2A3ɊVJ#DYSV*K|e)1agۚ^hc>~IN,@f 1(ZD?)LFieR4u,JEazt0#;o zP)]R +`/e ]޿9?|lKΦu:BoԙC_~K P+LATC?jZ;)L2X46 8Ã[23`lYwiXV0x ,*ėƹʣ_xXZq!/͎"GDYh%zZQI,xHq/2v F/!$Z^B}`>Ha[{Q>y zeiT4G4.I0$=E寱!/%Rj~8_PDH.lTr 6g3=z8grunu˰> JKĕKWlT$>:D%MX?of*d򗞢Nsq49&@`RڷϦz+#R8pF_OLyM0ۥE:/N<ׁGyNe=&j2˺^fp6)_VTF2<`>5X[t Tq?lOgϐ, $R36#s%ʅ{H» RacXw+YjKWG+}_yp$dFQ9=[hF,oJxfl.Sˬ#`T6 Z_)_[CdE(JAc_Jm`n_fG,'"j2IVR_oJJ&ӼN }`N{c`! yP!g#`B zc,1S`l2FdfVJt2. LlMMlfV Wu2s;=JUTD\IJ"x01=˻Ĥ@@@@@@[xG4V9/{ͥu"ؗ0(Uo[1fi-_˸_݃xEOXJ28wPFbXd tZe4`Zl\z&!:;_;̲SxePmR\qeV$,+JtRqkR0@@@@@DX@@:     i"pzKS(@@@"m퉒;Q_ d`}<Wڰ vxG6xQ§"lNyaqa$  A@<;^Ym&0#3yA2J[\RdA@<;`^ ;gQ7}[♧$c&C/4v%\Q  "씣X忥cb`]2G_/#ш/|vտdaG}e _y---m]-7!Gd0$mѸgϮ{vKA)!:0TP9忹vbzG 1l^x<޻_={B;֍c'H[E'l:[k~эjNGGpJ别h0~!C>7w]{=&%mgK?|Փ>gv][/>jhŇsQ?*,*3zc{u3gLJԭ̘AՏ>q&-_qÄI_#!=o,;oYV5QfgAj<1 &lHW OcĶ#E_TrD١SG SΛ@'κ~>=yO=7.uy>t}<(;EV^SOxڴi׌v7躾\ixE1aŤ.oܺPR`X%U N)p`4ll)޽'`7Ivïη^A\>`a>C>?WhFΪ(1պ%Қ:i/"t9;okiGrzVm(06Nu[ { y*iZ-bOW:H)W~Y-?zPpj >9ϰ9cdYj|m#YZ5uwhc}kGg$ICоݍ-;Bݻ7oԔXug~ܥҺREiEtpTZav5kSn/g֒^uUQ+?'5OxڂGܼ-XYG;?=Z_dz>m.;Zeckh ,Pp{n(-qfܨо~ZIF55%%%敁F> 644dgiw:lx#/7ח-Yb=vFXGrܳ}˶uxCPG;u۪~o۲#DY>_h "߆W_Ņ=A0(mƩUdt?aך;nѰeo{{[(Kc˓)°SX5*?}'^9﯐+orhbrM?qκOܠ,(wߦkվ:d׆uB 16[u~}μ3,[6|ѽWkJ5݃9}RBe jUϯYX_:O t>AC<}dO(G,6 v;#`j{v//egK'_%lYZZ  !ƶK6z'>OW00cl:YRղɥMRo^bO}&^JH$6F%g*'Ϥ>I ty) Ҋ5MM m4W^\\[(++44.T~ &X Q*93]۟'>gK~o_~)Ocl i۵뜊#"..+ER)˦MTO_ɳVwԡXn 2ʪDi iEC. mؽ+-Ε{VQLul׺h6o]TWɚ^} Ei5lXUѸ;Du˵ub~n|煏%tgsc">}z)袯iܶm_iOwdXM*R45kR[f/tuClc|O{Q޽_~wy l3>J\&~˯%ӦNRQ!7͘9F_7RjeSVTN-+{>?g?+~h$%=MKRT5CzF#U/)Yt-[gݶ=1$NAd*O84W? 5M>6Di:J:(Vby3eR}^=ڞmmks "-'YYs}3ct,DžMSFOY~)FVV6pORZ[m9aᯚ5u+蚤F$cΚ7_61Pk]ɩ%lhTu,IitWw՞^ɇt ?;G _hT͒5 kkڅuϙZRQ4 Y,>YzaM`Ǡ#,' QVmHS҅Qq 6SoW՛MTrQ@A_X%nm7;Ck{Ɨܙ 6Ohf иkӡ 'W?+%FZY0.|:gSOzۚ4?H99G3cFo پ}@WW8.u =REOGi痂47P߫&+KyԱ/­+O]C~ij\>k}cͪwjt6WWw|e^}-!tDE8KVO0b55/N3VH7d>3`lz]_VpU]-A3D XN9nY[>ZVKU %6<&`_9fmH3KU5tC>[g<1lq`6{j] c ]gze;f}<}I:QMVQ/o믟LGJRT#`KMˀ QM|v'Xtڳ꧎cHLukMb*;lR)EitiB~zp:I>(UNPy.KAdcѶV4lI#L$x~흖n :v0e͟Ɵ<|J12S_6y*8+c6+_6R%)J~x夯re)*ӳסr'?+w^y=IԐaCƍS߻k iƝR h~T(*~y°d!c}: Oby<KFm%qWk6u>:5*Oh+v7l-~_+OoZm 1H{@E'6ǖͨM)P2> WUͯ24+#9Эe΁͍g(b]Jook= e5ކ s`5h쫸 bȕkqr|iO $Ho f4E>;h>#Z]_PW&oKJiW#o^/DOA+b_-X)Gb0t a9%Rq۸iӗG-X1g~G̜=o*I=}!l4~NW^NCj~qEŔ^71ܡ憵/.E}8Դ;Gc_fcԯVYQ^6⳺uc nkkz;e2+Y,O?GctDM1~:r˜d9:B}z`#Lhp[{g"1j SjIe/?5rT66PhU_ [᫉ 2q1a'nԔLPWP.LKW}<(/'g܎`o^Ki&ED,DO ?xTjחm JC7G*Jc}oT.<3/Կs+{nj|6j5>]7\yy[B 1{I9u䌩#WĢ7DQ}qʔ#ތ:(? DZnW&" 2%~ؤDʥax*% "q0XiŵkEM]Vݰzau3`<~C<X6mWL6MQvfR6i5(Lmm&-@[EiWi"i &r•Wk+U^[3}U6V>SIVh7կfTVUM%_vECRV֭[nѺG'4+_s|چk;mʄi맕~n6f^5{̛f΢h4s,XҲ/>YdH7J6o{6EFG;µK[k6h15!+yԺ .ۺyݔ,;_y}GGO|//w[zS!^0:@_U cty+--9?SKMO-ND6J24e/J:Hxf+e'~ZدzKbYGg}{Mjƺ%5Oy H$Mbg 򙺬rlYX/Xg֜5+ӥӥ2i<Wt_HȪOg]%[PK| b+p4}\z_e60 tBHт$pμ.uϼf͙qmϘ=C^9~Y%ӓf솺QֶU_tPy͍. Kt~ه5Ĝ\ؕAIjuhq˲mG׋﷯ڸ}âKiК-_:kH4F`Y40aƣϮ.6Ćq((1ʱrv 0旅4kXH/D v9+s*%i{f֑GIt*wܼ^B!Q)Κ}?f6oED);>&0_KUM&i-.wOGW6N9O:|R9uCPYAJ+*e麿:Kp~No.Z/-󥋿ѩT֣U=L,, h#a}{8_#߲SW4{Bv̖ۚ{Zu#wGrf<Fuiuf ?_IR*jְFQ(GTW˓neĘL?IBT[gS=^9QE(H7e;ϖK҇&ʷ BQkQo34`@wO&UJ(~(&AuLݸT6MryjveӧWͯfE4:Nћ%T$8)8mMϼiz#3k9zlvy73΢hf yMLǜXg󎽝_{F~Cj}~wYƕg5Kj^XG7n4ih.15v CS8þazdr+_;c3n9cL~}QQY:]\ S|EM?z>M;bDCCKit 7OU7~bq{!6Pu={# ytog~D.֤0#o}+m$֪v4[H2h&&rL6>940}rfNFTgJK}ҵSϻF -OW}܄U˖^yTC-}ҙ,OvJ`hk(]xI/{x5cYw²m{C]! M 6XXzCdP=cQ)ij شWi_\m/Y۴,u*w):&4>h,v"Rmza?Cmҏ8"ttӛ#  `imfi|CX%PRKXMYPz!bucs-0²*)ON~h ]G{޷[^qZ"Ui^tF˦/w|i[7&CcELdєrjݻ-5Gm"gP)WO>Z3%(B*躡YzOgm uJ5S6ݲvm]=F$hRɖKąĹiצ+4d<ͭ3t5-^&Z\(2J#P7M}EN9[w9?}l(]UL&/kXhM0 hWWXV$`WRZySg_m{9]z҉s*${ 2e^W|?0j75Zraz&<q#oޫOnӸO>}?zYnXq"dNiFR_T/$}"6} ߮l=cQmO8 *cvՁP\ԍSTgk->|#CiUμ/=xyG^{g'oO :b3WV̦>@5U4k.,TOeUٳe̎Ico׋hhV|=Oߑ}Q|fK'sbLi2YRSOԭYߴ[g/E_S1|vvJET5-z覯$G}0Fg\trA.x')dQBzqw' ?Ybă2w+MmUlK yp|fyK00y$!"a#xцBl-,Xa0y `,p` j94zWͨMmڔN%;j>_Uw8]#'ndM#ܟQ6nsajJyUH"'U^5jQK3>7ҍ&\ռ㫏N%'t^)LزyͶ- 8ݴgKN<9]#J4 ּm㰡CV|gIݲ>;Q'1~-4t|E%>+ӶV/ZÏ<8hz\!D#e,rO| E_g>}y!m!ߛTP%Jv{4[_g|]a(|߯9w}?z{K,x]"UW^4 3`;t302w30@@%[0%MK}Q`:tS` ^$1\:[bx<`zV~WU%e$~#ڭffmgߎ ~S~>MͲ!C/+W*D\u-(iM5]xK;8oxgF==cye&k0= TliTNl}kҁ8ԣtC Y)֚paO\D7 X;vXMSQ5D_r8˃ Az{QIЫWw.@@ e,Ȼ4qj@Ag?QL?pï1 YQ\T@@@<#_rZQ<;`ݼ|x襝_p%9)1H5;"dM]@@Eg%)GK?W?Gu҈jinٲqcJV~=󓔓6m ?g/`u۲~]$ G]a CHW(B_iģ`]2uޣƎ/;\ js͋,C;XJzECYw]_nJ̆дX3I|R*_yvQmG 1l;\o%C@ |˫w*g=@=~D%#텻2<:ÇGҷMs kV,R"lm &/*Su(ڼu_{oT'?6LY2:W/$ul۹}ɝk޹t/6rZMQVIQo &M~q#Ou ]($γ`굕eӦV9/yqDwr]t|SD=+gt%eQsb_., dlΛPY!oM*KYRަ_:M./--_!݁eX'.v,+ 4,|+YJ.ߕs$z/6lڽ?IJ32Oq(VRУCۻYWg߁;ܱe!ϿnVS~QXCW2F~Ž}c1ػ5r|bUدmC]+ڹΝw}ε ~yhࡇ H7mj5`WvUؿ?;A2ioӎm,~KnTPG{󺵟]m>lRiҊs=ި}ifZU⮮p{GBޚˉ+Ծ}qc#=CMT1bŧ`o 4n vloXZ oNU ݽE'sʂ [e-!3,5PD3$s31S%KVe#hZ^jbӮ]j*F͡7573 o`ʥh~5,VW'RvqV'BLto|MIJ,)6sE^Hk{ݪ kFf1 czy^eM@/uu]M=;׳'CQ-ͭ;lk_+]F袙E3|TKSs]B  Yj@?( 6wmHWg[kX 姞9ζ}M;wZ3:}xBYD >7@LfTscMػ-IiHnz?PT`iܧ?&r2Tݾk[Cl۷o֭rܮ;3nV]p8YfGӂum;gG3Ag[Vz 5l {, vtx]}gu'|Ju/^,5# bʿ;q!ڡpC׮BC;::lA~IOgJ/p=gq\BxPsz2qt{ࡇik/3~zUWsSzjA'ϯ> %IoW..fڔlgw u5r5~F-_9OJ]M;&;D_5wQIi3@9Qh;tak P E(kj 6v4렧~t)KP^񹡐SΘhxܷk;E_RE%J.):~[9rsfRPp舑VB$>d?oHIÆh1d|)"Lv5Zv4ڶe;ݳ;@駦m{b^ gߢ]6߼|g_|]$DnrH]/՗fϤyҗ-5La*E_fW!==4oۨLDQ5c_vgkol"VGz_ryZ Dתٶ;FOR@KpwSpаJ^zRKeF [%>1,FIs1Za ɱSxJLmJY| Uڎf89_.LEF*'|/z{R +V"m6;wh/xo5BA™l˸(n~ 1>埴qm/>0Ӫh {lj@43RK&ynj*6|;uuu$lcW_>|/4<:[1kPLf"!&:". /\295O,,B>{ ?#3πM8A~g6ƾ_YV@W@(S f8W2x[4ӕ|b0 6hhòCC1 t3(^$zH/9C![nxkϖhsnNvoqI XEO[AaeeehWTHlYh=8ڣwk[_ sȐ_Ї!OHg@˾}9.!E;m/4g5iK&] (Q:lqG:|Hgg+w#MbLGvՙ꺪*ꫪRVI7UJWyՍPƻ(<`C+zdVRM8!S&N;ƌ,0S&֗VUL748*rZZ; zȘ64ܲajiz>HY^϶&Qc@i" y-H5U!)/}&Qesϲ,0Uuǟrh- KSF/ӄ}5fKbOSSӜ9s /p3k|KwΛV m^ W<>wna~!}B:aRmC[[S-!ě7YSwX' ҥ3?~u KrhHmc)3Q.K!͑>۶n;B9W՗D}AѮ_^Ӆ 腘TO%vo-P1vb[Xİ,}yv+XPTأgϜ<ژV#O,;l*7nVJ}LH}:xQlnܻ0KԞƔOOyikgW^;"?;0yŷ&՝\Rþ^}향&}Av"%;~xf*JR!dcCCÜ{u*mSϞ=nWWWyF}PM9ڷ~f-^SgF;7x \-FSyȀS*,\lA(fݷcK{Ӟ(}QAbB{_duāt#n+Yz9p _5DGC8"cjf}ME_Ҥ kX,9G9gayy9$Պ%npfs EJB\{bhBc}]VTD@kBh[0|̄h+-Js}n{aժA~jڏ~ssl}ꏰ㊙*|찯@Cdze蠥jz3GLjO$D-z} 6w&K"=<& F'So#?"߲\~5})N<2~;Au}ՔsWJc74pYCs3іS%K$U=ZU⸣>: (=1a 1G燻zDBXq}6ڡ̻lI&_/[^mճǯ{V~ W_v,uKR!꟞?aM4:G1،HMZa) _Ÿ\}]!&৿t#,KHo΃Ff>g„ 3WR F}˜Wѯo/ݽg߶ߟx}yӱa~}Ζ👙0 mWL6l>Pԫ?)1\KbzbCPgGqAϊFQ2PyXEjD"f,P2y GN)AHNhi#'r蠢9=`5I>WGy|*恡0͚NPO'S~: 5"1iN,s%~ʒS\k![(1Ս nLڰo%KnBۖFCXCїhj)k i5LtB)F}Fh5"F9CE[^O3'VCQ'I66RjCgj_P58?ZWczz >ճ=}t/+1ڙt1ʐO>{zС3?t) g-G#hg_"h_KL:b1ev*5 $O"xqq ~̬^s&UװKId3ϯ8dT[oʎ9Li5BL_+W[Pi_;9+4ׇBSb[gm, 49dJayX`.8W] K<32(a;z\H|ԨÇ4&Rum?#3ŞSg820zk%~Z& N^zl" 9r„`3ksnՃ.iH.imFz#ؽC )KCaTEApgcixw"99| [\ ,.Ʉ!l'#׺sH~']~~&NӻXuZ3rWnn^nJ]j.ϔkgSXeVHB"qǔ}qN^'㠷tuvnQNШśkWaٯW%F^Rˤ}uvCQ ;HKciz WsNd*}DLd2V{W<'pTGnxS>J!,./6?Y~ӧo5{Y.(s:Vȫ+Z9mB}/Eo446s~tPY L{K+a'2|SNbs_3nEjVGUy[Ho՞Cm >yʑuNZZPԽ!銩bڀ&t׬)-Xem׎ӧVu3~7hPfo}F޽zXΊÇ־{.zə|OQ+O7=ށ0'e4]I$69{~`擟o:r}! SǽbsEh;usjq#d԰bM][6RѤI\5,7v(< k ] /JGL0c _ L^Dl[jwnEMȣi 7S ux4L ɡ4\O vM`۾`zhnk૏))t>tWLOAk f^~u#'o X%be*Go=Dnm3~uv_{/ނkX^Dr,ٺڿ^~W 1YYiEӀTItF9;w>'$5=LK,x,>Yȳ4\*:#==$2rbI˛ȸo_?mJSnխnŻ`̻g&s"q&oO?[~򟎛}{^~Ӎ]WrvioЕ^qqɓ'5zd"z!&-N]"OFnw9)'GhJ)ޠ;nyFWݹ`}{w,Nux5kjn_yXV*oU>]gqݦزωg`B =c]΄CyR>|Ny2?~h%J?b; ֯~o\z+hKs.ӕj9G]^;mm>>П@5Ey]KEg}?kvJxRĻpE2[BLt#9%$[0%M2i/Y<#/<[!|']ar *@V@nQ     "LS}TGؾ6uj     ]d |:%fN,zE @@@@@ u(Km}e1TB/|B>dHwƯ&>rs,;dT~g5     e (AeFf@,~zHFؑOReD    D@0a;їOQHѰlj@@@@s 'D(RH2=)2<, sYyA4ї*AJ0 QHD !)f*%~$ |a!:,t`,R^Y xfxt6E" A@@@@2)c_l}e5]VzoF4.2B@f d&ˬY_:w1 (`u84N^Ĝ4B{|]ŗ]-`/Ki1     ݜ̙n-Ֆb=PE8lACbέZj8${Mdad42\65,hj2W-?fgjӼn5*4Eky 'Z}sɹ3y<QmػszHs9+|yNyrkggHlMHI)m)c k8~YxJCs?^j+Zs5<7[ȳyx,u ل$ P돺fHT4y//zM=⬑يĢ/5E1؀Ac3i̺ &C7'&+7'+5DuJb{vo ;`Jћ"ٴz֘ U*vY08+ ,~8Q,.Y;&`CWj۷A@ yJ$(H#,Wst97#a,#ةPܶE#"H8 wI!_W8| KRT<9`*:1N& ݁@} E:"a_p쁰pKjuP I>Zx#p{+Z~=΋&lbou-髻()6Un׻Vs^z Ƌgm52[yXХl[IV8"Ѩ}|9?w/|𡻣䣿x#K< `7}WC8SX""zYTJ`++ɗnhw[<~N赲鬑RwHErt-҃^4Es =7JCb4Q"Qo>.)H$ #_40󞻙_7l瑇, מK @*W8yHoՍ4q:ȁ}ϟh¢Th      `H;{̞?*7%%sӯoن4JG?M4 w}>́ME[nvw%-rjkk_[%BH:θ "LXuӼ?\I!׫OC%;~h ^x0?/Es䩆uDiXL/ZH'pdrt0$@"у`E$ؒ-H s`?P@@@@@@!Z~# wE]S HHM~;pH@@@@@@@E{H]hW{HWg;}]W^3'uԈ+n` @@@@@x#|Jy'{Rhܠ')99<      pj7/OmW}vkJhmw>_?s|?`0      %#~z7Nsc1s>nQ?rX^FNmoxE}xٚEkW۟חu:Vj      hJz{i_ܣQ/,.Y;&`" mgǶ]=}6o_0J[vъ>릵ǧ:ព qQt3W%Rԕ\w&*uRid&z  z0Xzm9 b[#jf;": Ov_ `ϔSy /_MLgؾf}$Jפ9Jbym999nB#޽,~9,E<#, Mb:H_,-NH `w '2U~V.W`#WMMf&-E='*QgqR#yOx5Aa/@Erihw dN#>3WM| 4RU(cuC`MIiPJ%v zΉ8Gz[_&\~(@oe04G=c2,,cr]tZ??OLN&/"?S l:53pLctٖրaD IɰΒR!h0ݰ'IIf8zp ?+g9㖅<*MvVu_G}䆨(ݑ- l~ߤ;ێ3A|QnQJFN2y'XC$#' GFH`*Lv|RgWT@~[tĦSD˖" wUR3lWHjzgǢ܃E!4F4@nEXrcy3Mf8_NDB7lOse0]4A)x? R7 }2dp& s@@d[|ўr扵Tm3m .W7){TPL0 ,jG$4)YbȬb]$I3ȪNRmjL =V0==aAN]Jku V<@,>av^g(q&1L(;IdR?#Յ.0\P iUÃVFe{]uǥdL/k5  u<\< +ڒeG"_O3). +7K%(q"~U/ #=mƺq&D `Tnny 9!eA9 6 \`ozH[ BAe@sTD<EɈ5*LAkH3^1[.HiiK$kP%@@ ފ-9! B:KH      BQ[Eqj4Ax)g`y0g%u\V     IF苓)! RV8Tv"%@ }٪)e)"` .f[ 0;YTfxP0[پJ%1fA[@@@@@@+0X=A@@@@@@pw]xYaQ6}TAF`%.\VR/V[+>#RH"FN,hl*`r[<ཱིx-iS[[[^^r*Eҙwͭ\IgrjugM2 ^*Iv_g ftnDX7r6L,LߣYjR>xuFїU`P~e}M%$,o^fL=e A@@@@0OA@@@@@@ bLg0J&!N7S3`jTpW{\Cϰ@@@@@8<} Cc/q4K:a!+їKۭH     "R9Sp(OR)KYH]QmtWHuF_Ĝ)oaf҄\Jzg @@@@@D#0~~C.k+)ȗh`+BpTFIH m= C5 } ^!_ SEk/a|o _afKd'Hm x0KAH`I,e<'!o(c_Ⱥ[RiN:ɇ4%O`2ǼlHnv[5lݹwgo6IYm A%,szlT3Eбd;̦v+|m¶hSU@'&$Ɉu%/kB@W~LН'~U}C XT_z ]DP o 5~JGj {$}kx<]Tę}؋rU}E '`X[WZW$z7y㼀@MEv/P3#Aj~<H?^ߙ=zrb$`epΡZ{1:.\~U # RyȲ]!WxV fJ_q",7KO@hԒG}f,eK7[ "u?eTXhL`[Te,KTd"OL64 MDmTm-{NlCD@ DROdS9osc+qۛ^1!h dOA[HUK_ٻZ[XX4ͣ VX:ģ6t,oVCY$B=QBv}iR$ dRޤh'L[F ۻ87m8[Mi  #p'EdQ)̊mazWbW&c"d5' (ыy #QW0,$KL)BJYN_X[8&sKwVÏ)ARM+KY&gԅ̟2v\ao$3I9.I I)fknlT U􅘽+O5yLVqڅA[ UO˔zʫo*})e"`+WbV*C|-U_P5ӖdoƲ~يA, p`>gi[j8$Qj DEWa L+hRbVDZ) ټDI#5BhxH}PL$tKgY2Dpt k4q @@@@@ ?weEy.z9S]Q      F\Zm^RHQڑ?r")qÍME[nv˞w%-rjkk_[%BH:θ "LXuӼ?\I!׫OC%,V<     Ѝ FΆ     % O(ɏ+Ց!ub:N|brsV4&S}" MIv5oƱ A 34d/=v'bqХNo&^_UN$w"gݙuFπ̫خL7 `OOI|9j%EPn:kxmg ~V"ӄ-&wMVi0$TO<3h熒 8\3Fna VDE{:2&jKE-^]yywQmWk"f$2YbZ;Es i=Y#Sd, -      )2Et;]xYaQ}TAF`%0?'OKϣ##$D"R m=$!E"bXcxE"GXekk [OWj#c+i\gpk Egw;A8 X`ؒm2Ë{A.~=ƨ'Qymb"pw%-+ks:AK:0=k3`U~= ivV[.E= IdWBM*3 2 涙~5T#%і!xvp2I%x7M_5=; ^gy&:=+a5ԩd ,k!      A2Et/]xYaQv}TAF`%0?%Rl(ȟ`DI98"xv-7e;Gܒ~9/׭[$gnt&Mi^q$ӐU dwua bf+JFu#gT/b&]髲YuzIsюӼR" w36/(4y.RᓞVuƐ[6L,_pX}aY􅆘&N4vvzr&+z:@2H"G_; Ӧ:ꗥɮԗz5( %'eQ5CY(pD{/N_XF_,h[ec+z:@Eym999RRD|H6Z#/"I9+\T8ΘN5Dj n[]Sr[.HD,CLuc+q2Z"{+5 2ΰx$ p}"k;ih(@D&` >M@bTRp! Wx%ʯL2V~__>W\Ck<9^ 9`.[W[V_yēPȓ%miKXk³Ćd>Óڛ|A|aR}!f'r=0T޺HA@: x&c!R?*C"`XN)!:OBй\J m]*A&'/`&e</ΦXh1e4$ETR8ʼQ;CXgrpd4 ݇tZP,"<='Ia_8LsH'ml*`r[?ཱིx-iS[[[^^r*Eҙwͭ\IgrjugM2 ^*Iv_g ftnDX7r6L,daj髲SutS4տKy] aOTl5PmaJhXjN NkzJ c uq][|YJHi#4ΖK@E`ʦ5u#R]Dy `0ڕ9zN}%",Y_ K,Ȇ7q*nz苧#}cl9WZ$fᷔ]ڒYK4J]פ  `_ilR0zCh6>MD +Uڴr t׳9M["agc3h/3JnN,HGOT޺c^9Zܕpkn#=5aӸ]iBb  '` H@D}WEbqw Wm6.i/2'[^xi@@ u` sS ~Z12NSm8yHU~\QV-83OR2E4V+5m^`xA@9Ul52 %0CetW. 7'{&WM"Md(!o"^0v=*;u/حuH NN0D_bN&_yö$d E6%VqlSLu@| ũ@/K=Pfj=] ~Vsğiy?BchՆ³l5s~,# :0ܚ~cUa+~~(e@!`/c/Ի3y s'*c PUEfAa!UjAr 7$? {BzAV, '`W 5j 2tFղ ~-Ee$/WGEkYh"qK iN:Y64[/N_Ev^7Cmg:JiKQDV^JM@ X`       ``&{~EYNq9S"!xqK }}y{IaOEH$,oiIRGK9Rľ́ME[nvw%-rjkk_[%BH:θ "LXuӼ?\I!׫OC%J )̊E],*E#H X)x(@@@@@@֍ SA@@@@2K:cR>z .7 etCc敚91%wBҋ^Fy-g&X7%_l&Z Bް5oJz"QI/y5=ksM@LX`K*0ͯ?9!W`0ڕ9re4"|}yo1lB/ oXP͋0JSqu& KIE?YKpGm'+ͫeWmtgyHU=O+:S2IgIeLP$RsN_{?e5nnp2ɠ (@@dfXF-\= R]~Y Fjo+cɨRMgI6g^a/9}!Y2oYw~ A@ ;X?Sy LMLO( /=Թ[iE~QOePvxρQ4mQDXdw^T*+D@%$SK3`&2=>^ e2Bj*ʋ-¶ŕ&=n3͎ !jmU~*>5^l-[G ᬊH A02X"ww1F~r<|"z"83^mCjwo}$YCJ5@ /3.l 06|LpCݙV{YjC^Qy[t@[d(FO%qP^W"[=[@0E8҉R=Z}+S47 0Df<ҙ&gOTzV~j:ctvòLP [8+uh;"7 :cJq@z@%` :o@-A[=? \4s( @@@@@@@4Ǎ.7tᲶ|"O GhD~H*B)7$ >x/tƊD/vYRP:@204oaN$yyT]7*3cKgff$WHkgF[zupEwێ( a0 E_ipH%bЩsF*B 6ZqKT0V{d[z&JlKHɛ7XE]Yyv{뉇z'EZV&_!@ >="!lobfg @@@@@D#0~~C.k+)ȗh`+BpTFH! ;9?<vnκ_NmmmyyyuDIgr7r[$:nk54zՂi$~`يA@@@@`0@@@@@ U-m'zӎoѫ9B_՟,HYEU%klj&5G?0m,+<̤I[YBMF3z5 ֖TbIHL}1 `ʖ 25b FZOLO[TM}Qkjȯ } "ב*iіQފ?#F_ ɻދ] RfE_i~۲N)C72ՊEhzE*=޷.Epɂ ֝[c##-f>HH7ӃlȺ٪Mlj9zs 9}aY%][[q@| (xqF,g4e2Ji鞘 oN@1!\.$-G]Шzv_xE(Jf/idR -\H  5l` K'g|*U_qwq -W]O/RZ]ᶀ[ޏPhfj3=ESE_ح"Kv;OCoz7׊n^EEh&LG²bݓ{F@P)kxEd 1$*IW47)"pܱ#|3w HWb-[6Uؕgi4Pj yo|q`BzuMVvMYiUG |1 !- 9TAEkȯh#̯چ'Ѭ/rUnvY>Cw8RQy`4Qf$&BL`G-bEBmhrAi+SJjo"+Da,Q11 WzDFY @F4QV¬Mta'kgwlQkkk.㘶A6чWDUnqʠ!w>]ns ˂f6A6ܡ9e(_% @ _8"ya{(ECٛNKPxغ@@@@@IpLs-^VXd|cu>g     ?ntI,m/)ϥacFD"R m얅w+[-BIf\B e5PYʒ̾P@8UFS(@E  Wϲxb`K+,kY"5IO  ]#f؝D^ƞRLMGqCٲN{]K+NQk(#+oK=qIjl@~HiKUrBm2=-UL {->EF4QE餩NMx2T/AJn+-$SXy!z9l_ %~C ݍYFWB,4I ug2֥Fzg"\YiT놻% pQ4'<-\dA%$JglNQ,+$I(Qn+-ؒ^Q+zxݪ- NerzZ1pވI<,S^̖kgS`Lp^ࡊ4 ݐ,_\-`'^?[Jqry*F">}+ku*4<9S}ps6r%rUK3zЭdYl(cA2",HF+OBBଢ&Y]q%7]YV*װu hkjj*4tH\.Gh+ Bs8( q5vLP$SR{9Ő5H p >kb 1\M8;Û^2ͻI ij7yvX<;CH7.9j < $K߄oCy} Vekry{L}V( L6S(u#/)p54  P[C%yO+|w=*ҟrSubHy-ʭRm!GRTxw&C"`mʼи949}!Fj3wPT5 ]=F!@Ah`5;du'X*T~ξul,- Y,uCV*?&k|PJMԻ/hsB3Fuz~&;^gΩ6j·s"jD9zU k* $K<SCj\2"ss`E xynD_1s_D=9 MQk|W;f1B&}{F#fRU;TgcQs}yĭ1/{n2}'s1OXo!p QעCo a'#gAQ^-zyR 4r[P @@%O @@@@@@ $`V0B 7_if}Y]h\A@@@@@=g͜]wN+ {AɴCwXX5Ҷn␱"|0AL|/[i,1[ZniD&eMðQ7bI -&enX wX579W('N&[/XU[gu@l#$@@ 0턆օȻƃgIչ*4Yo?kϖP @Q&H[X%-*"3`:sD''nipUYu}4J{S;h dlC( ŲY |w{MvVC2@ KPY@H:ڝL'V^MOWOTußO$$$[gQ6^>\S qTV0!Ixf_& דjq8sEQ g{kdW9r+=ATuXy(Vog8>o\g5&w(=,y%nɫ<2phqntbX_sKیK n0jɷXLnJϘjPy#՜ڍiץ?jGʹ>9JO-v9D/&sMhhM4q( oJtP@ H n0u!fk;;g~_b &yCm4wq: j`{W2ԉtȮj][%WLLnpQe7qX¹EcpԱ0 @@@@@$,Oh @@@@@ 7_ifТ4nfϚ9yԫo;e\P,7Uo_0P*JH<vmom6t-{VY|Z[ޞZH٪H:]7ut]2Z!HKp bm4$` l      P[_>-inzʢ@ڂjP̙w-yC !sytڀ~TpSReblGWilR:r Ҝ3q 1K~Wٗ\1K|gA}+0YG,)?1CKr-ԖzfV8\Imk srV,9sy=]]mA&sa}EsINR␪Zԓ-ZcpJss9p:P-,EQTd^&C&@+h&^H7R)O$~,z5 *LsBiAy>\J.@H7 Z! oJIW<ᴢ!mVBϰlC'5lL|+:BkbDYQծƮ 5,lv`_y{L=D@Cmɺ%3$,,6Vsh1/ Q;ѕ8"YtOTKF5X0d_v;qBj<^4vmɇV\ tA]1o e$?*OI 5BeΤdj^u !I}q둸B|B:Vq rL$5Iy3fQp,ƹ]QǮ 3:`^8(ˊ[|7ͩ:Ƥ)ZɭJ)g['7bv-|` b =y1kBE%iyzM( @ #qQr!< lSrqXuZCS7 nHV:?l5'B jOI ^A(l5Hȇ@:`B̾f] x Ta!x:Q+Is gGh' gNQKgTZLԮ2C )LN)ItFBk8DVgzD?Xk1m@1(pH`(nXn--%и, {fϚ9yԍ[L8@[eϛ*FޓS°GaL-s'6m4;s[Zr/=vmmL'qU mL'u1Vn5d\-CwMZh H0@@@@@4 1u [Lɗ=162DZqP>T1ᡡ<~-U\av E إ:z?^`>.qvnC 'X XU~mX)UC'ȿ/ڃy4f5^?] -`9~̳/C+8F!z7tp@BAKr[Y6Jb-B;T5 l'ݰI0=\=ðQ>ÊPVucZHAgn2m2PnNk\ [7!5aEQl)0&Z\]U@@J&`U  :ϰ3b   Q&Q ݙ)YNhRJvY)l% [W̜IgV7*ECjs ŢU]i"O`A+  0d_Wn %?(r:N-POSO y@  h0d_)K5=[H'U9pUd29LR1$ey]ӢPs/9&@2"Y wா?ppO >O ́kKjVےHT_<-ʚ uU߲r5^*sJ.u"{9⇉e&&JL1(+sM( P[q뀩k.0ߞS/*// j!Z_j.l,~}8L0l >5SáIDZԈܢjqee$P @ %       `0+!@@@@@֯۴yKKkq.4.      ٳfN2u='+VJ#T0QS܉MM\Ėū֬\b-inooϿ][-B[$tl[$qծpM.C-C $%`A@@@@r6L-B|+堽l.tg6TOݣ.hɾ\o0lQs#gCyD˾9 iI.&L:&ʀ42\rXyרsK7 4N劥YؚVx(LiA:D&v=4=rF@!gn*gғ쌹w%NLa_n}8"=(Q]:vHy A@ 4 | , Dն}& ֭%  Ŝ abjq~E=cf9@uDL:2Q/ DB`YpH%rMָ`Qbh@u̗}`ZQU ̮_7t%KƉ!g犙?qv"]~Iba^c9hd ٗpS\ "%OE(哚i1<MD uhm'fn-JPP gӢw5 @@n4CA/mƒw5B&ٞUO|4dZlKgNL/Fźm,R Q!\UB5L< ĭFz]|LݓK8ԫnKnncKmzǼzn%],`5EV]qjFm ^kp7x?s_\S7]d0q;=䚸e@jK n0uek;sP]&&],uCBK&j+A=!suM` G6˰E xz!{`GHrJ a675PPvC֡P^4 `%V$`ȉ@qu6oiimcZQ      Y3'Oq˞hbLy[H{r\()e}[ۦ&`g.]b˞UkV._֖tww߮-:nj-?jMs]&]̡s0܂XC     ("M~чfhk4E3\U|9l     BrկU-E(s"]on>[#9|HC8Ƕ6fr0' ꩚WvFunwbe/Qlb n[Y]˗n}kqe+;p'% u-T(>c*( f_QhY{4=VO(mYEØwÕ^ Y[9tkUMaU}OZcgvH >$iC: :4iǫ G7ɟ9Zci/c(0ML9[ݰa YŋP~4db  Ids䉟uZ[gCnjW4&VGni L%`W |m<>AyKn~il7/KQdդúJпNo+rtF|)bG'72?p%~'$OjqO[ ]lb2Vb5iɭpV&CKk&Ӄu qIp!ThX̗}wӋ/c{9"N:g Ms )Ð?OM̕2`bNgv"5H"dau#(gQ @:`m/' yGp(gd~p1AT7DLrgK(z: (U7@.j ݄@[ 16#^s2 U/^y憨dڼg:_:wmI>kc[^yn<ZO/pή+^S!@ Sq뀩]ɮ| P)C !xGW)Q:7[Qɩƻ)&Yj|B}5凎D$ġnB@% ]H4$Pܰ~ݦ[ZZ-BB @@@@@=k)S{-'GJ<# År0<'6m4;s[Zr/=vmmL'qU mL'u1Vn5d\-CwMZh H0@@@@@4 1u`,.Т̡[W0W8QB@WœU<͌ $bЮjZLvZhv)抛D5'kMXFIDAT+Vؔ jΨnHOyP7/3W+($4~ȾhV3Qgʘ`:s:HJ@,!8RQanW֙ ЄV)=-՚& 0W;Q<1 !/BG$͎k`)lW2]c! @ qQWnP>ፈ49M>ik2婿w.C3^q6S/8H sp%f]HN0C/ &S߷!sp ۵=uM* 5!]ܙ4kdkN X)xZﬡCEۥ N7mmȹwb~nf_RXh@@BZڒ =9NjCj^2).[kX!O@57@͵2znVs:pw5C|L\?cN %3Z=>⬛-4X5[w5l]@"& >oOD&EuU$`MU1h@0xY2@H9"4ت6sd&9wd j0藨>& ^{z0,7eL(U܎8hU|eggS薢JΝŧ!7 ^}|%2}0ynOմj!i.5 :%4ൃuS/ \ƪkV @(M˾jѭC,$_$7 1tFq9&~& ]s msSx i׊vyz!j:٢JCa ʀЬV.Uг7CjRL=Wa 3ܫ:|Ԏ9iIn4*~&JM&>#2j89;(3X^%ன =}.S>!=S/9%T|(o3`FZ.JُŸ? % Ѐ 'yr-Bq)j/:HdׯG'ҳt 5-HD n0uek;zM|r|W ϡV¡䠌yxN)'; P2ʖ H4?[= >jwE=Sr1l :hwH0*Nڴu9" @uc 555VHyB\.4c~(I׉@`B G@ 'Pܰ~ݦ[ZZ-:؅Ap#0{S^}-㊅bP.} RT;ā7ϝo{k۴̥Klٳj+ڒ۵"E2OV-E2Xiפ9r=AR7][kxh@@@@@ k gT,,WZ9h.# ݙ2I(vWIuxj/eRp`7nԹrDhMbez!@ -Ĭ.%r-"`,Ĭ.! dH se_yPI}q !y],m%}q&8 A% 9294) `j;pMZFыkE>}L\ 5@$0a+OJtz]?+D>6I`N>0Ja'h ճ(*]kkP50Z  @R&`9kڜ2tz+-'I&#R@KZEup8KQ@ꀀQz J|':WT+Xb@:}ĭ'j'*R&3/D07&I^Q +u(<}``E@*݊q$}j4ehŸ?kcsku@@cn]3rb_ vB DZkRyɚڨ/EEf 5i(ji(k`hf0o`~~jWsKĶ?lF)f\[=Jk K [梪ԹZCo+hy߹<&QB%VN<g\g{yÛ+t!MtP @ĭ.,rS/>+Rl]%:37\.U̕tф?[$ P&xܢVEbR5jG};bHY Yп#V!@@@@@t:L 6ml.( @@@@@@ٳfN2NnW,˅roW( R0R,y}[ۦ&`g.]b˞UkV._֖tww߮-:nj-?jMs]&]̡s0܂XC      D X9Ԗf!f25R8Xm9ZS[-ǖ:=>_ #*z<;HV) fA&?~~#;MHdZP Fy72nVHLi:+d >F+w-C(|(4 3ջ@ H k %  y0Jģ_Vπ\{c[:WQmeK<@'n1o^C\II=Z7ѿnJa j0=5ve!Ӱ\9 }&40[Cov4cԎ5d O n0ue9A].e~@],:&v(oO9F&B(*IκJ;5 Uڀ\:< LJ:+Ef0UbFkBqb =jh@ K8Ca(H      6ml.( @@@@@@ٳfN2NnW,˅roW( R0R,y}[ۦ&`g.]b˞UkV._֖tww߮-:nj-?jMs]&]̡s0܂XC      D X9Ԗf!fҷ2Xڋ|AMuΐ²=88  sm evB'8k}(myF]34]1=3LȎ@B_wopgvʉcf\13S5 @8gI##BthGN׭[[NDK0J1x:rQǸ삼zn%-PV$c!@ gq 1Jhs `8`hT-6ZV@8Ţ!Akׁ]Zxưj @ XV1~P8babNV̞\Cr"esi,KUPoWpMݹԆ@LܔXtwf9+huug]wJ}V#Z8EZu`\,'`G+_!ߙmxP=E`F꟧ϼϚ7)iR&x橕PӘj*+qBH>e@}$*]ÁtHD@Kȕ/HL 7yʭFk/㰷շ p |y\++[S) I|W(uyR+nu]Υu?]YIIy2j+ξ0n^2g͹y"=ZuP5ku &@DiCs'!1p!@3 ~97`~cnL%2>  ` 7hVHsKâ3vs%s6nHآ[!Y =3̾Ԕu7^su$М7Ls% ksQV4{!<{V0@^C,^AwjۨzMj2IotQ+!b[q&E6(qӹ$O-nAT\ڇ6.3`1W%_+>G,<YP.J1:YQ#RuNV<׺&]%! @8p0@f @@@@@@Z mڼZIJ.( @@@@@@ٳfN2u='+VJ#T0QS܉MM\Ė=֬\b-inooϿ][-B[$tl[$qծpM.C-C $%`A@@@@r6L-L]˧%T[:8G 2K]K͸}9f(      @dFW~"1:t$~ ٗ[     |h0.>>&     $`10<娃6oA:4 s_@@@@@aqw,     Ľ^]/;h     [Յ}f!sP@@@@@< P@@@@@1 k ?J֯۴yKKkEeТ4nfϚ9yԍ[L8@[eϛ*F#\()e}[ۦ&`g.]b˞UkV._֖tww߮-:nj-?jMs]&]̡s0܂XC      D X9Ԗ@\& Y[n6$w~ =Yk&TטGB;z]16d/-Ƭ d AD&`hџ%tg}1wd7ۯT߬8[P7#5w^tpΝOԮ]O<6ݳ~i&gԏx-SW3`9OڄzLRzՔ7窚&Pׄ@&Q&Q ZSy??j}-o9|Oq:jGg]vBN[}!# f @pݡz  A}AKn?gt\rw9O9kx `N@qߎi.%A@@@qE;?N_8rƇ^|GfC+/} Ll$`{Eʾ|+I 5ss mb&tgn*!OW@5^C,ɟiۉ)K8z6Z@ B ZՂjE>]xGN 0u!faG|w~Cz"/d :kUN1Cņv.HShkD tY(82͏>&jE X D&`h~%;mSa%Vi.5EB ǬBn<ŌP;uhI?*¨e=t d_Bme_5Yx2A D&`,7Y<4CˢƔ)Nٞ|"V8_W[Vco7i>2_&EרQH0!Q1ie@@πib8aggUz<)eԳck"9u#MVSn׀nk,* Q J@'ԹН Hk1ǥesܗ3`s1:ԚA/+rumi@2M@ gL>%Н9gs2ibثo|g Q%4i@9sS|p6uh]zR ZjŒ I9M@}-UQqRP{#= r"`>RWɰ:j#O8S^`@UKX̌sji^?xܝk@@x$"@n28,Z9(&@@k7"ևV{ |U4꾝[CP, %]hk d[#HU\<3*Rd!Ic`/L5HdL܂h] F ,`DFb&Lg?i0n4Xl`W<(hX_=xikoi˝nͭe7ݸeU7_yο>uQkՅk5lTp(*L:mm4SZ:vN_˳ݽ \r @@@@@@  RT* šHiX*KH8kd@@@@@L 0]S**PV.0(     qhkx@]kGh}x쫸aM ."B@@@@@@'ٳfN2޺2}JSԋ|;+?S˘c_3qwֶi ؙKزj՚W%-9kEod:9Zod:9u\I%szhAH腇[Ɲ}yؓe\O ~{C9X<Dl: +C_O.||/ק?wJ5xӠ q҈vw e_/e_ 0u0YAIݤA@@@@@r@/ޠGs0ϟ2\,x[#2J]L͵`D L뀍Ñ}N\,6y{hLQ傽@@@@@jF6z󡗌 JF_I_.P=%bZfơaD\.ʥTldp-{yC>%8.      Whladhoih }, 6KMC/{ӫ.kf^r̀}NjqG>QlGO;&tˡ*.|QaBJ-Oɣ/}@ '0vgҝ_fB@y UFSsHsn' Ѵpn8,{D,޿4e~ O[zbMN_rRL@ן߷Smn~7>2d& 8%$]Dw45K$b.%A?h\aEOlM4G oLtzKsԏE5ؠ&MS#7&eHZJVBO>(  N(@X*c CeGoU. y-qCoQ.y;VU M1E:c;,,,fasO bpwO9CY;4P>j`OUkXLV7ъ&B!"ghK|J&26y:%&䖔 J$D`DMOӥS@@ ikT.sޱ#y{эUI7%!eЫRnTs'ܡ-6K`}5 [QkÜEQA !ûnxx3{.W =7?QDΤ1!kNwDNMDHꝇi9NAO/Y芞9  Pghʋ .dȾϝ=Fy6sޣb#rT(KMl&F2K"ẻQ߿Q0'#ipUsHϫYnC'"^󢆾,i“|JD$ Q$S&P3fНRN/:$\Y謑ȎDDAbZ=q@a{]ы7/_ag?{zu`4KV@` -M"ύ T>%]u;e4'`T珤դ"k5$Ҥg,V%%0HZIs'CeDdXX-F:=S*  ЀCzʋI_:eBmVMy0Eg^b#L6lf/l |$纘POq =?q+'7LPJ?@'`Itz'-5 &ɘ]\wn193` 6$ .fZ5zjQz@@xJ*88%bC~7#V{*ŽiFxw>,6lzE BԢ( x+Y.<7q96J\O?w }ÅYcD-TjT=i奆^لН8ʺj $5$oR4b  U,iϜvBW6+ bSsBA!@H<<0Mӎ滾'`?SFG?/8YIDN? fQo7,&BjtYЊQ;RmB  zw bD_3O6{}V/=H=^&7XiMw{{O{_[]v14ʒ }+>SM0y4'Ѱa1)J{W?{fLӷ_(ަE߳U1BKFU=R4o  0'nš 3SSV{y_CRZ {}b wh#YlY,f_0ѿdK7mO@ij xWshe Z7}qw_:7E },֯۴yKN^V$ȼc}SR2=9Pv}J # q^d˄qMcT=M޿4ExiRzb4tZ.ο>qDaYm_`Fwf:D߱[Ca {=s.:4 &,};7~&X;nֶi6ogBK/_6@ Ec C#C{KCGKCšAʻKMC/{A3Y:h!craxXd#m!!Y\t^6.&MyHL1UoN΀Y𾅟e ${Ӣ @M|Ih̿_ |^VE_d%/jVըTX!!Zl 3=iі.ȐLd#Cg`hsdH=<1Y} GS;hb@`jʁx:~1}B^[+o^ި/SD#3ꂦ`F#0N')~ F>z#p$obwda=kH C ęt"0:W~C9TM;/c᭣Ff @@@G`4{+PwG|׏kMF̙HBi$0z 3z~y?-|1gO7ŠÅx-xCu75s^2Dwx_6ebLw_:~0Vˮyͬi"眥_oϛ6 7㉑+,+%  I`t숗&6ЗkλcG-[o#}^<'%;%|ïfe ɟ-omi?]o}g" J ~;6jkcɧ~T+_'J‡?K.SOy_/YYN>!BӵwO*Ĥs`Ooq߻uV=۽s-SyA"k{w7>dc-j%ƿv}=ꍟ9_7/F-~{_ɰ 'Sw ?yCǃ1-!hڗ>=6Q䕀9C_xH얯}n}2G M:mŪƞ'Yh&VĬc5q[ڵ-|1o}kj+hͼN80e^e=͙{W,ZtrXW̘Wڻ(@S@AohimkO]>gBA)D6zPu}WaGN 鰫./^G̾::oFZw=dJ__ߎs4noQ-Km o{˻~qMT2U^ó~+m}}q0}NyY\eVd_3aKMwp   N,M)f~Oq3դ즭=?aӂ\haaT_d2TnGͣqSW5O^~3uHL |{HtE΅2bQJVJ*şGL?0o{arO.}UN 4q4}ufeeK.;{o,}rK_$mߠ񚴨g[ONJ nk% ʠb +4{]}آysCt.Q񝏵PTH" Q]:QzgD)@?Q*]Zve/X9E_Ewԉ;wѹ^jA<5:.}}̑w9979/+7i{T>]>,>q1dU&brYW|u^s֗>~F/30*UW_uTȮkԑP=JZpX%z=p0RLW0VQ+6!_>2=AUD9o7/\"ҕoK/6)|TŨ7mMbT8V/.;wﭦ=4Uڱ6qJmo7W}uMWLыR7hޗXpZwz}R;C/O|Zz03fgfS!c>۲s/^b/xo~|_v+}7I+ɗ0eou/{ 7'}}>Ml 15zmr'g}^_#Q b$P@@%`tU>h mܼF:bſ'UkW;xWR/Ӿetתּg/K_䯪mCɨ:P`!0%&ygQisW<_}/UM"GVx'­z㏿{(h⋞*ls7G$$QzJ_ӣfǼ#y 𹎉^^NQBnsoxKL޾o}tֻ?KNRҽ2dJE]dEe_;:`O[ӳui[c};'Qt{NC1%eg,hoG=c'90om]tW!J\Z%11BA  MKbc3s@lbԿN}{6^.>m9 igN?/^W3S_Q)F"ʤ"3֝7,94v2{-}>o?rcD~omQ*FCKizk靇mٹӱGLDN%?yxNr2Nȃ?o|b,:kgk'4DIMѫ8z.1^e_"IE22Eigт҃N5s'>{e[{e_vֳw=V|+ͷ\b@ï{<+Nك7=ߡp|ӧ?|靇ޯFUy2ඎ9QF=7߾kәK&=f?ޞ.{3a}"T@|䤶I_wWn'xrk.e!2Et1^zQAcƲ^|Xp](Qe'л:zvl֣?twLX/,>玮~{z^~N?0E\eށG ҴiD{}G c}O:&h%pwoۺWzӺN>j;Mƍ[ k^,yiՉQ/8u^2;GSK7f5Lr]ZpΟ]}ү`yw橿]>jt?b6 ,l @@@G`4:z'MUR]ży8J\u%t>G'W]wR۞[miwmk,;_u|Do]LLDo~zCו:y!/5}{Ը箿1Ǿ#Q~͢˻F_ZXs ;;W/;{M^77{}ǂE. q%`.^#_w#0{^?<9G_]Kdz`{;^ED K)ū;~G:7ji w{oXDVZ==mBB;i5Y抛thIJiF&%Uצ7%jGHw6JzϹӾ#C-C'+^}O.5 :?uUC3uyz|}45o񂞎/."9߿AaݛG|_w a!3b  #0Mx֐(ys_ GדOܧkwr6/8w|e7h!!EBfvQ+YM=/H+jk{[{}_8t1/;:wVb;w=o]Q4kD3`֝(Tݻ."_X} Lu|?~j1U@?;/_{=wZQ=y]2%TN[nmd/{y"ة̃U21o*$s6szT+'|!12~q~V2ʙ(xi#J\ïmw '>D_?o{HZ&͘9kBJJ7,0q]@}Ξ=mmm=id_$uXu 0CP( : XssUeG0ɵ /XA7ǻ_Yz\ݔ]qksjm p=;^tˎG }S>7~ o~uhYzFGǼe.ٯn[(\! b͚3ĒJU5j`{F \}s+*CMz7L~|ioLhK+gzm__UW~o{N>S3REʾ<?q`1ꖏx+?Q/|Ze~wF#30$`P @tF3r|/QǔA`?7]O.u/|K3:ع{sYkIT^?;Q^fc4$2$z6ݺ~O^qWXzd|+\% )Tk|6п?\QYH>H hwI]I  n8R7s' X2_"K A@ hT@@8;Wzd 9b?,㐀%  Kb'j[pWx!Kb$`x40r'G o;J̥ X2#K A@ "?  .3q_3/*p !(gtCAJCrW+y:Q唢‹5R+[FƞbIDhfx@KQF U bj7ƘrF[~x#C2% !Hidq=ϑ!Hd3L]^УeXQfォ9(  OCqW @@@@@@&[ @@@@@@3`h><=pl> ` &"+͇Ⱥpv%W) p "N5@@@@ʋyqa"D94 EjWuJJR_p ">@@@@@@& سzi}҂U      PFܪ      ~!3`bza"       ނQ7r}Q     0J(} NeeA@@@@@ S! ڢK j)obYf@8' Ou9%R)ww-4F Ĩp$zK7C{L`D6z4      l 0lSނ(-ݔ(r0_1Lu<ȟ[W -@(s*,hX><=pl> ` &"+͇Ⱥpv%W) Ő1tP@@@@@ 5o%,j@W/{      @q3P*@}DIENDB`Dd D 0  # A"nhxAV'lrD+ @=nhxAV'lr}a;7͖"xYklTE>guFF4KShmIS * GC %ǰt^akP c{ Y~ְd+0ɰ93ˁ@+lj񝫍@TCxzy-h, ?dXBGXXgΧ]Cӹ"0.u_`.:Ǎx>"?G5InѹxNoy< (涍C2ܢ3 1qd]tN[t <}t8Eз xئk7!123#izv~ŭ~Z}!-N~UOkITԮڇUF|J}Wwe+\bqb DײMԯ/xb*BrooXou?˂_X&W ~K5JyKh.1IS-aXVy98 Tg)]B}s/X%9W6 BugGc3<,| 9QǰWv8 Gxf[P6{% }TK,|T;QjmcO<;W27ޯ^)|9=E@9PKTkaeYC'J&7G|4tq)}f5rN:D6QNMkzmB|K5m7c{/|M܁@Nu :]#A qrKV9.p'c_ $~{={g &wO'hgdہ~z}3lJc}8|@ '6f];]lv62=E[և^m_XsZ_q\GjYvg:e8bb3 #Yiui /^['T)|4/D}\Y˕Sev&:IǦVL>Dd b  c $A? ?3"`?"J<[`d@=\J<[` 8b'*xZMh]E>ޗܼ1M\DgH5V_SPߋ`+jSqEDIB|H %tօua-E.kPQF7wB~sNN4ͣ3|s33}t0&:_"*?9ȣq.Zx|s2=hhͫyt{,=広μDZWj~[JjFz:H+H{ݬhu?.MM̳{6mzț}quBԑ]zJnI[=}ӥ ;t6g^O<3^nnE Ɲ`lu o tsg}xm߲}J"ʦ.h(}|G.r#委/+/*WJ>J^rد`_k))O+i%V{ߑ0]zƜ^S|}?SkG1߃Bƚ5PΏo%3Wo+yN:*zZnI6,GB:Fz . k#=>ud =F7\SٳAVsVV?:iU?gcVMu\$;Yqs~=fXހ=m/L/]a@eÌ/~dKk D)d/u_cN $\(_V$v5)4qRc2~'W: 2= cѻ(o&PAyX`.G)gO(;V迗}[k7zrZ+V9&{Ze(sj_> |} VVL៍δ'W൓oj_캆c$U`׆eSESs9`o`c⵲ w&ҭ8OfWO׿9Li?ldeW3 Nro'sr(N.cmrZYYrs%|Nα6/=fÚrr\QW+:ٿs_cV֗c QL hx-okBCXvXc 6c~8ÏCG ?W!wW~/'[٦ ? &Y/تs&royښ<63 ؊q1^*3'sYT\<cƔ 08TMt+:QAu`S|WNr/zO\ ?z@G~LuhFT?>61&g f1rPR|<߰=ž֝{*Ôdۃ:\~H <לܩ|kB)u`T1{`Y*Us\+Zw':̅~{T:t;ۧmr?ӱ[^tFtCWU"*w4>nmTnV*  !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~      !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQSTUVWXY^rbcdefghijklmnopquvwxyz{|}~Root Entry  F@n[`Data RLvWordDocument ObjectPooltZ@n[_1066897713!FtZtZOle EPRINTa8!CompObjb  !FMicrosoft Excel ChartBiff8Excel.Chart.89q  Oh+'0@H`x Michael MarshxMichael MarshxMicrosoft Excel@fNj@O jX>-( EMF8!` N F, EMF+@,,F\PEMF+"@ @ $@ !@ 0@?@ !" !" !% " !O  &% % +>% % % " !  O&% ( % '% V0W|tW|W|ttW&% (   W% 6|W6|t6t6W&% ( W% 6tt6t66f6f 6 6W6Wt6|t6t6tJ6Jt|6|t% % % " ! >% % % " ! M% % % " ! M~'% ( '% ( &% (   % % *% % % " !  M~% % *i|s% % % " !  M~% % *@SJ% % % " !  M~% % *  % % % " !  M~% % *@SJ% % % " !  M~% % % " !  M% % % " !  O% % % " !  >% % % " !  Ww% % % " !  W}t&% (   % '% 6% % % " ! Ww% % % " ! O% % % " ! O% % % " ! O% % % " ! =   % RLArialIh Ih0Ih@IhPh`IkhpI_hIShIGhI;hI/hI#hIhI hIhIh Ih Ih0 Ih@ IhP II% TCUUAUUACLty = 109.23x - 8.135h0      TTUUAUUALPRhRLArialhP I7h` I+hp Ih IIh Ih Ih Ih Ih Ih Ih Ih Ih Ih0 Ih@ IhP Iwh` Ikhp I_h IShh% TTUUAUUALP2% TmUUAUUA L` = 0.9595h   % % % " !O% % % " !O % % TTZUUAUUAZLP0hpTX0UUAUUALP50T`}UUAUUA}LT100T`}L|UUAUUA}LLT150T`}!UUAUUA}LT200T`}UUAUUA}LT250xT`}=mUUAUUA}=LT300% % % " !O% % % " !O % % TTUUAUUALP0ETT "UUAUUA LP1PTT>UUUAUUA>LP20TTpUUAUUApLP3% % % " !O% % % " ! U % RL Arialtu uXIu 1#^]U|E S]VW3} wd=s1ȃ t*It;t!EPS0]E 3ȁt tIO E ;C}E% TNUUAUUALpAdvertising (000$)   % % % " !O% % % " !3w % RL ArialPS-]E 3ȁt tIj ` VC} ta[j Y}t9EPS0-t`u EuuPuhuPuE 5u EuuPuhu uuSuh2uP% T<vUUAUUA<LpSales (000 units);    % ( % % % " !O% % % " !O&% ( '% (   % % +>% ( % ( % % % " O!  " !  ( ( ( " F4(EMF+*@$??FEMF+@ ObjInfo Workbookt SummaryInformation( DocumentSummaryInformation80 FA\p Michael Marsh Ba=u =hi9X@"1Arial1Arial1Arial1Arial1.Times New Roman1Arial1Arial1Arial1Arial1Arial1Arial1Arial1Arial1Arial"$"#,##0_);\("$"#,##0\)!"$"#,##0_);[Red]\("$"#,##0\)""$"#,##0.00_);\("$"#,##0.00\)'""$"#,##0.00_);[Red]\("$"#,##0.00\)7*2_("$"* #,##0_);_("$"* \(#,##0\);_("$"* "-"_);_(@_).))_(* #,##0_);_(* \(#,##0\);_(* "-"_);_(@_)?,:_("$"* #,##0.00_);_("$"* \(#,##0.00\);_("$"* "-"??_);_(@_)6+1_(* #,##0.00_);_(* \(#,##0.00\);_(* "-"??_);_(@_)"Yes";"Yes";"No""True";"True";"False""On";"On";"Off"],[$ -2]\ #,##0.00_);[Red]\([$ -2]\ #,##0.00\)                + ) , *  " " `nChart1Sheet1Sheet2Sheet3TZR3 A@@  &monthsales advertising >  FAMHP DeskJet 895C Series Printer: d,,HP DeskJet 895C Series PrinterLPT1 ,,"d,,??3` O `  ` O  ` O #?3d23 M NM4 3QQ ;Q ;Q3_ M NM   d4E4 3QQQQ3_ ' NM  MM<4JK4D $% M 3O& Q4$% M 3O& Q4FAQ [ 3Og 3*#M43*#M4%  i`M3Ok& Q (Advertising (000$)'4% tjMZ3Of& Q &Sales (000 units)'4523  NM43d" 3_ M NM  MM<444% (2@M13O\~-P  Q'44e@?ffffff??@ep@]@d@@Y@ j@e>  FA  dMbP?_*+%MHP DeskJet 895C Series Printer: d,,HP DeskJet 895C Series PrinterLPT1 ,,"d,,??U;;;;;   ?@p@@@`@]@@a@d@@?@Y@@@ j@.d*(  j  0NMM?U ]`L-  FA"L-??3` ` O ` O  ` O  3d23 M NM4 3QQ ;Q ;Q3_ M NM   d4E4 3QQQQ3_ ' NM  MM<4JK4D $% M 3O&Q4$% M 3O& Q4FA6 3Op  3*#M43*#M4% v SM3Ok&Q (Advertising (000$)'4% wMZ3Of&Q &Sales (000 units)'4523  NM43d" 3_ M NM  MM<444% l@M13O\?*P Q'44eee >@7 FA  dMbP?_*+%"??nU>@7 FA  dMbP?_*+%"??nU>@7  ՜.+,0 PXh px homeChA Sheet1Sheet2Sheet3Chart1  WorksheetsChartsOh+'0 ,8 X d p |MA6*4v?IߢݪAT+ &O&r/+\CF+g\}}g):V6ܭ_f{{{6={ց=o{{6==;Kw={{ l{{q31Dٗm c`^[^`^?{=Z z8kY6S`^^^{=Zj{׌-O|}{طط n```;6}׫ghO1ΧgxwX<>5ϚKkBҎr‡`l-E?Q8ݜ'<,#ϲ#q``|7tKǛ&ckO2@A11iINm^mҾ pBS%Jp:_oUEk 1}9^)ݫp:SyTG'g- Ѥ9<9G|2GR8)}Gt܊/88>Hg@)ݥp:.{p=~ˤIl owLtܝ/8ߧQJ?4TO)=H|죔q$PJW*0 1pþ-9ʳ/SzipM1p&Õޯp:p]B9;wS_#~`d<#NSmb3<*~W_  N)^)zy_ Oqxdb끧h)nySN)j >W`_ c 'Y{wl (>f*?Cgl>xK_αc'Hed-T~d| `?EDwD| q`g9sF֟_I5$[c1}G/id}%Tn$Y )LadT^EU6C,/ؙ,pH`1]}{_$ґVPy5V| 1`gހc yEd9Gdl>xv_1,ʭ$kb}"ϛ/D[JHǰN{rk"7Pyv| k|"`oSD@f5| Z|>H|pIl1;ORzz#iM`g7?,O a/اJ*=jMTn!Yǰf̓jV,G גlǰ;#R7;{ؕ]ٞg^_ qrZ)j]y#RwĮ<e]ybW'f/HJ+ĮÂ] Nؕ!veG8窑ؕrG8 _2Q8?sf5#İ.ҏfB?͑}İK%fQx?cWzbW^_dhĮ,ؕ|_|(1,1+ vQY:R$ Ce's<$v#%&dĄT!&78++ !vŧ>U8+< :8vq]Y9,}z^wq9}_0]5DN o|߹)QS3\c%%ݤ?HzDKzTO%"ICIK$$-8IH:0HKzC$=Aҡ :>i^uI|_ dfN>YXUo:MdwMW a^E7iϧ_QAͫ蟕zA@_ͫDF ?˹{ #_>dA:=Ǩz*WG y9MFĠN' - ΐko;`uѻ[=wwz2&V]:Yz-.;@c^&0_H ԍshf}G= 4#8K]t7*Us%1*2LO?dͱJ-?h0'3kf/W(wұ7Fĕzoͽ"V2z#+-+:Cr./+M1xկscGɴt M:UԗyS컐J1 Ҫk 75^*f3o6| (~D?ZZmߵ^E~s#Aޝ' &L!嚷uL^|#2y!ObIL~db@(kyN;4f2^y[^LMLONTi|{ ͅ߄{Iס;[&Jlk<>{t+ƨol=Β~vwaU Jş@.ϳq,~ ǧc~?ief:f}}D}k}_d}ٷfgU5͈sB8%~f0{e~{?f白71{/b7fuo0{2{3{o`^}0wce~ʚ/ؘ~ g[fU^ ZM"3o,E}|> kY93y(wa_rZz!߿'9+Q3׀_uW (KiXAKѲ :KѲХ1t{uQ3oM^5+$YhكnT![ӹ_>-06נo&G(V-hi6@ fv' X3Ly.$ XLspKA_%%/ ˆW, "3t94_,' : W, U8QC< ~ :^+?eTVȾ\eFc? JŠy%j2= yb,5q2^+l'Mr<| XVmVC;ͱOzώѺ8jGAjx!~M5qM@!/3b<.v |[b+}O~#d:~J1Nݑ+8+('#"]yC8s np[[KY8wlx)n턵cv28Cx{ s>?_U+8N/=+:wkd|NY~|޽++ ΡdXS֟<9w 4~fH^X.+8P.<ʉP,GE4,S,w(1*S Nzt( J'h|+ zug#`&Ǚ9:.W^qX+8?~;'gNr]y ω\8Nʹc84s\DxwL~H$#hY njbb6-EurS]^A$4v)p(X+xCdWȗ#&t|dX~+؛ߏZނQZ,oq۩?ߥM5ovYUSW N,Qt]uWOwwE˧wm-ܮ%$#_SW,:u @0užSp{1Ӛ.b_πL]o*\bS>v vSWWgۉ8Y-L]oU\GS95bbg5A::p&t:Ð6užr['^՛b+,>^Rl{ctf|r3dM]\^F9d+bskol+͚++Wt_yV6Bh}_ ;t7숩+-+i~o&͘Qɱ]VSzCfd#-`ױ]UM.ՐUb߮] u%*x]h'J*M]l/7޺ݗ"u~1oɴ: F~g>+:kg!jN_At&޴&:\-$0yD-D/"~1]Bd/=rv]~ 8]|Y zF&ߴUl:vr;pE:T1XZ$_MNCHhO5uZT,ҿ(@O?OПI+K~]i=/|Tl<cL&;|H?K6::s<뼇ޡ> B)tjO}D"s,ew!2@4@ ېޮxs uRaumI(?};1sϤ6ƗEi~ Ő\W$߻_CjX+^2X7J"Gކszm66k[MH+"HDڂFMm%ֿj1RiB J JXTŇj"' "\=s;g;gg̜Yywh.meaf"Jy21+:kM@'% 뮪栅yWb;_#7W |_5Rm Wf=bݫc Wd@:7 -/<)tY趶JDx;o#_G.|p+޳m.#dZ>x6K{Qbm"Y,xW{BTiwwa r9N~F(O)+%ߢJޫ%%/ z}&|$2|Zg|V;MPީSrk_Y~};n~q?_ĎZ`|w1`~Gjb=QcȾٗD(Qu(xQ?>S~sV/+r GiQly]Oo\Uxǵ'p]' >y71=Q~H;F~Yi} -v>8xƶfZ$5ΤW#^G|(c9-vx1i3ab緈7#Rk3_8q'Q?N΂+kkZ,}[N=^OMƏEP1RZ܆ex݉Nν_#LƔ Zq׾IZtr_^3";èc#D#J#P#Q#S#W#[#\#b#h#n#o#q#u#y#z#################################$ $$$$$$!$'$-$3$4$7$;$?$C$I$O$U$V$Y$]$a$e$k$q$w$x${$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$%%% %%%%!%"%%%)%-%1%7%=%C%D%G%K%O%S%Y%_%e%f%i%m%q%u%{%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%& &&&&&&&%&+&1&2&5&9&=&A&G&M&S&T&W&[&_&c&i&o&u&v&y&{&&&&&&&&&&&l)* *c*c,d,f,g,h,,,..Q/R///011(2)2.2<2N2]2w2x2{2~22222222222222222222222222222222222223 333333 3&3'3)3,32373=3>3@3C3I3N3T3U3X3[3a3f3l3m3p3s3y3~333333333333333333333333333333333333333-4.4/4R4S4j5k5s5z55555555555555555555555555555555555566 666666Y6Z6p7q76878?8F8M8T8[8c8d8f8g8l8m8n8s8x8}88888888888888888888888888888888888888888888888888889999999999999999999: : : : :*:+:|:}:::<<W=X=BBB B!B"B#B$B&B'B(B)B*B+BBBBBDDFFGFyGzGGGGGGGGGGGGGGGGGGGGGGGGGHHH H H HHHHHHH!H"H%H)H*H+HHHRISITIUIVIJ JJJ J!J)J,J/J2J3Jg<<+0TVxY*,,,%,=,W,q,,,,,,-;-Y-x----- .).G.f....../5/T/u//2w@@@@@@@A&A=ATAlAAAAAAAA/BCCCCCDFFlFFFFFFFFGGGG HR .=M      ' 6 G [ m    <$<2<@<K<<,./123456789:;<=>?@ABCDEFGHIJKLMNOPQRSUWXYZ[\]^_`abcdefghijklmnopqrstuvwyz{|}~<-M:  '!!l ,b$p] {X*J@@(  BB  3 Z  S  \  S %uLA"`B S  ?88$BM_iltgt&4 4(," 4(s 4( 4( 4(ܥ"0@0@BBM4@4@BBMB*urn:schemas-microsoft-com:office:smarttagscountry-region9*urn:schemas-microsoft-com:office:smarttagsplace MMMMMMMMMMMMMMM ))++/,4,./0/00J=L=*B+BBB+C?CF+FFFMMMMMMMMMMMMMMM333333333333N"r"x"&W)l)/,d,)23k56889+:B'B(B(B*B+BjDDG+HIJ]MMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMMM .dfCS\[x"}""""""### # # ################$#%#&#,#2#3#5#9#=#>#D#J#P#Q#S#W#[#\#b#h#n#o#q#u#y#z#################################$ $$$$$$!$'$-$3$4$7$;$?$C$I$O$U$V$Y$]$a$e$k$q$w$x${$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$%%% %%%%!%"%%%)%-%1%7%=%C%D%G%K%O%S%Y%_%e%f%i%m%q%u%{%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%& &&&&&&&%&+&1&2&5&9&=&A&G&M&S&T&W&[&_&c&i&o&u&v&y&{&&&&&&&)2.2<2N2]2w2x2{2~22222222222222222222222222222222222223 333333 3&3'3)3,32373=3>3@3C3I3N3T3U3X3[3a3f3l3m3p3s3y3~333333333333333333333333333333333333333-4.4k5s5z55555555555555555555555555555555555566 66678?8F8M8T8[8c8d8f8g8l8m8n8s8x8}888888888888888888888888888888888888888888888999999999999999999: : :GGGGGGGGGGGGGGGGGGGGGGGHHH H H HHHHHHH!H"H%H)H*HJ JJJ J!J)J,J/J2J3J?@AD[H[I[MM`@`*`X@`2`6`p@`:`<`|@`L`@`P`@`@``@`x@UnknownGz Times New Roman5Symbol3& z Arial"1hm[DB[[ B' B'Y4d}M}M2HX?S\2Module 5  Forecasting Michael MarshComputer Services