Scribd Logo
Upload

Welcome to Scribd!

  • 04 - Forecasting

    Uploaded by

    Thu Hoài
    13 views136 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    13 views136 pages

    04 - Forecasting

    Uploaded by

    Thu Hoài

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    of 136
    | cel HOA SEN UNIVERSITY Ac8s? "OPERATIONS MANAGMENT \@ r@ 1 7 SS PART |. INTRODUCTION TO OPERATIONS /, MANAGEMENT NS HOAN ecragasng Dr. PHAN VO MINH THANG Forecasting Provides a Competitive Advantage for Disney * Global portfolio includes parks in Hong Kong, Paris, Tokyo, Orlando,and Anaheim - Revenues are derived from people — how 4 many visitors and how they spend their ae money iY 2 y)° Daily management report contains only the forecast and actual attendance at each park Advantage for « Disney generates daily, weekly, monthly, annual, and 5-year forecasts : » ¢ Forecast used by labor management, maintenance, operations, nanny and park scheduling Vu §)- Forecast used to adjust opening times, rides, shows, staffing levels, and guests admitted Forecasting Provides a Competitive Advantage for Disney 20% of customers come from outside the USA Economic model includes gross domestic product, reae NG tates, arrivals into the USA A staff of 35 analysts and 70 field people survey 1 million park guests, employees, and travel professionals each year Forecasting Provides a Competitive Advantage for Disney Inputs to the forecasting model include airline specials, Federal Reserve policies, Wall Street trends, vacation/holiday schedules for 3,000 » school districts around the world Average forecast efror for the 5-year forecast is 5% Eid Average forecast error for annual forecasts is between 0% and 3% Dees. Objectives | SEER 3. NOU DD . Understand the three time horizons and which models apply for each use . Explain when to use each of the four qualitative models Apply the naive, moving average, exponential smoothing, and trend methods . Compute three measures of forecast accurac . Develop seasonal indices . Conduct a regression and correlation an . Use a tracking signal =H eet Session Contents | SEER i Forecasting e of Forecasting rs ic. Importance of Forecasting 4.3. How to do the Forcasting Task 4.4. Qualitative Methods 4.5. Time-Series Forecasting Methods 4.6. Associative Forecasting Methods 4.7. Monitoring and Controlling Forecasts = A What is Forecasting? | SEER * Process of predicting a future event « Underlying basis of all business decisions ¢ Production * Inventory - Personnel « Facilities =H AN 198585, The Realities! | SEER + Forecasts are seldom perfect, unpredictable outside factors may impact the forecast + Most techniques assume an underlying stability in the system + Product family and aggregated forecasts are more accurate than-individual product forecasts = Dees. Focasting Areas | SEER 1. Economic forecasts « Address business cycle — inflation rate, money supply, housing starts, etc. 2. Technological forecasts + Predict rate of technological progress « Impacts development of new products 3. Demand forecasts « Predict sales of existing products and services = Dees. Forecasting Time Ranges | SEER 1. Short-range forecast + Up to 1 year, generally less than 3 months « Purchasing, job scheduling, workforce levels, job assignments, production levels 2. Medium-range forecast * 3 months to 3 years + Sales and production planning, budgeting 3. Long-range forecast + 3+ years « New product planning, facility location, research and development = Mie. Distinguishing Differences | SEER + Medium/long range forecasts deal with * more comprehensive issues and * support management decisions regarding planning and products, plants and processes + Short-term forecasting * usually employs different methodologies than longer-term forecasting * tend to be more accurate than longer-term forecasts =H eet Session Contents | SEER | 4. Forecasting 4.1.The Nature of aoe tance of Forecasting 4.3. How todo the Forcasting Task 4.4. Qualitative Methods 4.5. Time-Series Forecasting Methods 4.6. Associative Forecasting Methods 4.7. Monitoring and Controlling Forecasts es Strategic Importance of =a Forecasting + Supply-Chain Management - Good supplier relations, advantages in product innovation, cost and speed to market + Human Resources = Hiring, training, laying off workers + Capacity — Capacity shortages can result in undependable delivery, loss of customers, loss of market share = Dees. Forecasting in the Service Sector | SEER + Presents unusual challenges ¢ Special need for short term records ¢ Needs differ greatly as function of industry and product ¢ Holidays and other calendar events ¢ Unusual events "Mis Fast Food Restaurant Forecast a - = Ss . 2 15% |- Qo = 2 10% |- a a S a 5 5% o a $ e 8 3 oO 11-12 1-2 3-4 5-6 7-8 9-10 = 2-3 4-5 6-7 8-9 10-11 (Lunchtime) (Dinnertime) Hour of day =H MN... FedEx Call Center Forecast | SEER =H eet Session Contents | SEER | 4. Forecasting 4.1. The Nature of Forecasting 4.2. The Strategic, Importance of Forecasting othe/Fo Feasting Task 4.4. Qualitative Methods 4.5. Time-Series Forecasting Methods 4.6. Associative Forecasting Methods 4.7. Monitoring and Controlling Forecasts = Mies. Seven Steps in Forecasting | SEER 1. Determine the use of the forecast 2. Select the items to be forecasted 3. Determine the time horizon of the forecast 4. Select the forecasting model(s) 5. Gather the data needed to make the forecast 6. Make the forecast 7. Validate and implement results = Dees. Forecasting Approaches | SEER « Used when situation is vague and little data exist + New products ¢ New technology + Involves intuition, experience * e.g., forecasting sales on Internet = Dees. Forecasting Approaches | SEER © BeVE einem i earereky e Used when situation is ‘stable’ and historical data exist * Existing products * Current technology « Involves mathematical techniques * e.g., forecasting sales of color televisions = A 12822, Influence of Product Life Cycle | SEER Introduction — Growth — Maturity — Decline « Introduction and growth require longer forecasts than maturity and decline + As product passes through life cycle, forecasts are useful in projecting + Staffing levels « Inventory levels + Factory capacity Product Life Cycle Introduction Maturity beret Cost control critical Best period to increase market share Practical to change price or quality image Poor time to change image, price, or ‘quality Strengthen niche R&D engineering is, critical Gompetitive costs become critical Defend market position. Drive-through restaurants Internet search engines Xbox 360 Boeing 787 eee estes! Sales players 3D printers Electric vehicles rae = Ds Product Life Cycle Introduction Maturity beret ‘Standardization Fewer product Ichanges, more minor changes Optimum capacity Increasing stability of process Long production runs Product design and ldevelopment critical Frequent product and process design Ichanges [Short production runs High production costs Limited models [Attention to quality Cost minimization Overcapacity in the industry Prune line to eliminate items not returning good margin Competitive product improvements and loptions serene EEC een ase Ste rae HOA SEN eet Session Contents | SEER | 4. Forecasting 4.1. The Nature of Forecasting 4.2. The Strategic Importance of Forecasting 4.3. How to do the Forcasting Task sa: cuatatveet 4.5. Time-Series Forecasting Methods 4.6. Associative Forecasting Methods 4.7. Monitoring and Controlling Forecasts = | SEER 1. Jury of executive opinion « Pool opinions of high-level experts, sometimes augment by statistical models 2. Delphi method « Panel of experts, queried iteratively 3. Sales force composite + Estimates from individual salespersons are reviewed for reasonableness, then aggregated 4. Market Survey ¢ Ask the customer Ds Jury of Executive Opinion | SEER Involves small group of high-level experts and managers Group estimates demand by working together Combines managerial experience with statistical models Relatively quick ‘Group-think’ disadvantage = Ah eee Delphi Method | SEER + Iterative group process, continues until consensus is reached ' « 3 types of participants * Decision makers ° Staff + Respondents (A = Dees. Sales Force Composite | SEER + Each salesperson projects his or her sales + Combined at district and national levels * Sales reps know customers’ wants « May be overly optimistic = Ds Market Survey | SEER + Ask customers about purchasing plans « Useful for demand and product design and planning + What consumers say, and what they actually do may be different ¢ May be overly optimistic HOA SEN eet Session Contents | SEER | 4. Forecasting 4.1. The Nature of Forecasting 4.2. The Strategic, Importance of Forecasting 4.3. How to ue the Forcasting Task 4.6. Associative ein Methods 4.7. Monitoring and Controlling Forecasts AN 1288. Overview o Quanlitative Methods | SEER a 1 Naive approach 2 Moving averages Time-series 3 Exponential smoothing models 4 Trend projection . . Associative 5 Linear regression model Pau anata = Mie. Time-Series Forecasting | SEER * Set of evenly spaced numerical data « Obtained by observing response variable at regular time periods + Forecast based only on past values, no other variables important « Assumes that factors influencing past and present will continue influence in future = MS — Time-Series Components / Cyclical v as HOA SEN UNI | SEER Components of Demand ‘Trend component —__———_ > Seasonal peaks AL 7 Actual demand line x Average demand cover 4 years Demand for product or service ‘Random variation Peet Dees. Trend Component | SEER + Persistent, overall upward or downward pattern « Changes due to population, technology, age, culture, etc. * Typically several years duration = ADS 19398. Seasonal Component Ds noasen ccu ¢ Regular pattern of up and down fluctuations - Due to weather, customs, etc. ¢ Occurs within a single year PERIOD “SEASON” | NUMBER OF “SEASONS” LENGTH LENGTH Vat! Week Day 7 Month Week 4-45 Month Day 28 — 31 Year Quarter 4 Year Month 12 Year Week 52 eee) | ie ee Cyclical Component | SEER + Repeating up and down movements « Affected by business cycle, political, and economic factors ¢ Multiple years duration ¢ Often causal or associative relationships = Dees. Random Component | SEER + Erratic, unsystematic, ‘residual’ fluctuations ¢ Due to random variation or unforeseen events ¢ Short duration and nonrepeating =H eet Session Contents | SEER \ 4. Forecasting 4.5. Time-Series Forecasting Methods 4.5.2. Simple Average Method 4.5.3. Moving Average Method 4.5.4. Exponential Smoothing 4.5.5. Trend Projection 4.5.6. Linear Regression = Dees. Naive Approach | SEER « Assumes demand in next period is the same as demand in most recent period + eg., If January sales were 68, * then February sales will be 68 « Sometimes cost effective and efficient + Can be good starting point =H eet Session Contents | SEER \ 4. Forecasting | 4.5. Time-Series Forecasting Methods | 4.5.1. Naive Approach 4 | rife rage Method 4.5.3. Moving Average Method 4.5.4. Exponential Smoothing 4.5.5. Trend Projection 4.5.6. Linear Regression = Dees. A Sample Problem | SEER + Let's consider a company selling electric equipment. This firm recorded their revenue in the 4 recent months as 580, 640, 720 & 540 million VND. « Estimate revenue for this firm in the next coming month: 580 + 640 + 720 + 540 ¢ Answer: Fs = —= SC = 620 million VND Simple Average Method =H eet Session Contents | SEER \ 4. Forecasting | 4.5. Time-Series Forecasting Methods | 4.5.1. Naive Approach 4.5.2. Simple Average Method ing Average Method 4.5.4. Exponential Smoothing 4.5.5. Trend Projection 4.5.6. Linear Regression = Dees. Moving Average Method | SEER + MAis aseries of arithmetic means - Used if little or no trend ¢ Used often for smoothing + Provides overall impression of data over time S'demand in previous n periods Moving average = Moving Average Example Te) ata] pea LUNE] R-7 ty MOVING AVERAGE January 10 February 12 March 13 April 16 | (10 + 12 + 13/3 = 11 4, May 19 (12 #13 + 16)/3 = 13 4, June 23 (13 + 16 + 19)/3 = 16 July 26 (16 + 19 + 23)/3 = 19 "/, August 30 (19 + 23 + 26)/3 = 22 2/, September 28 (23 + 26 + 30)/3 = 26 "/, October 18 (29 + 30 + 28)/3 = 28 November 16 (30 + 28 + 18)/3 = 25 "/, December 14 (28 + 18 + 16)/3 = 20 2/, = Dees. Weighted Moving Average | SEER ¢ Used when some trend might be present ¢ Older data usually less important ¢ Weights based on experience and intuition waa _ > (Weight for period n)(Demand in period n)) average S\Weights Weighted Moving Average 3-MONTH WEIGHTED MOVING LV) mae SHED SALES| dacs January 10 February 12 March 8B =e. April 16 [(3 x 13) + (2x 12) + (10)/6 = 12 1/, May WEIGHTS APPLIED PERIOD June Last month July Z Two months ago August 1 Three months ago Septem 6 Sum of the weights Octobe: Foregast for this mont Novem! 3x Sales last mo. + 2 x Sales 2 mos. ago #1 x Sales 3 mos. ago Deceml Sum of the weights Weighted Moving Average eit ACTI 3-MONTH WEIGHTED MOVING )a Sey Vat) dacs January 10 February 12: March 1 April 16 [(3 x 13) + (2 x 12) + (10)/6 = 12 Vg May 19 [(3 « 16) + (2x 13) + (12)/6 = 14 1/, June 230 (x19) + (2x 16) + (13/6 = 17 July 26 [(3 x 23) + (2 x 19) + (16)]/6 = 20 1/, August 30 [(3 x 26) + (2 x 23) + (19)/6 = 23 5/, September 28 [(3 x 30) + (2 x 26) + (23)/6 = 27 ‘/y October 18 [(3 x 28) + (2 x 30) + (26)//6 = 28 ‘/, November 16 [(3 x 18) + (2 x 28) + (30)/6 = 23 "/4 December 14 [(3 x 16) + (2 x 18) + (28)//6 = 18 2/, Zz Potential Problems With A Hoses . Moving Average UNIVERSITY * Increasing n smooths the forecast but makes it less sensitive to changes + Does not forecast trends well + Requires extensive historical data = Dees. Graph of Moving Averages | SEER Weighted moving average 30 25 20 Actual sales | \ Moving average Sales demand a J FMAM J J AS ON D Month Peaaen =H eet Session Contents | SEER \ 4. Forecasting | 4.5. Time-Series Forecasting Methods | 4.5.1. Naive Approach 4.5.2. Simple Average Method 4.5.3. Mowigae Arar Method ial Smoothing 4. 5.5. Trend Projection 4.5.6. Linear Regression = WN is82. Exponential Smoothing | SEER ¢ Form of weighted moving average + Weights decline exponentially ¢ Most recent data weighted most + Requires smoothing constant (a) + Ranges from 0 to 1 ¢ Subjectively chosen + Involves little record keeping of past data = WN i882. Exponential Smoothing | SEER New forecast = Last period’s forecast + o (Last period’s actual demand — Last period’s forecast) Fy = Fyiy + (Ay. 4 - Fy_4) where F, = new forecast F,_, = previous period’s forecast a = smoothing (or weighting) constant (0 < « 1) At_1 = previous period's actual demand = Dees. Exponential Smoothing Example | SEER In January, a car dealer predicted February demand for 142 Ford Mustangs. Actual February demand was 153 autos. Using a smoothing constant chosen by management of c = .20, the dealer wants to forecast March demand using the exponential smoothing model. Predicted demand =142 d Mustangs Actual demand € 153 Smoothing constant a = .20 New forecast = 142 + .2(153 - 142) =142+2.2 a 12828... Effect of Smoothing Constants * Smoothing constant generally - As a increases, older values become less significant HOA SEN tRierey Impact of Different a | SEER Actual demand Demand 3 a = Dees. Choosing a | SEER * The objective is to obtain the most accurate forecast no matter the technique « We generally do this by selecting the model that gives us the lowest forecast error Forecast error = Actual demand — Forecast value =A,-F, = M8... Common Measures of Error | SEER « Mean Absolute Deviation (MAD) MaDe Actual - Forecast| | + Mean Squared Error (MSE) 2 "ee. Sores errors ) | « Mean Absolute Percent Error (MAPE) ” 3'100|Actual, - Forecast,|/ Actual, MAPE =20 ais Determining the MAD | SEER erty Parte FORECAST WITH «= .10 Bee he UNLOADED ; 1 180 175.00 175.00 2 168 175.50 = 175.00 +.10(180 — 175) 177.50 3 159 174.75 = 175.50 + .10(168 — 175.50) 172.15 4 175 173.18 = 174.75 + .10(189— 174.75) 165.88 5 190 173.36 = 173.18 + .10(175 — 173.18) 170.44 6 205 175.02 = 173.36 + .10(190 — 173.36) 180.22 7 180 178.02 = 175.02 + .10(205 — 175.02) 192.61 8 182 178.22 = 178.02 + .10(180 — 178.02) 186.30 9 2 178.59 = 178.22 + .10(182 — 178.22) 184.15 rae ais Determining the MAD | SEER ACTUAL | FORECAST | ABSOLUTE | FORECAST | ABSOLUTE QUARTER | TONNAGE TG) Peay rn Patan UNLOADED SS eee a=.50 Rot ee 1 180 175.00 5.00 175.00 5.00 2 168 175.50 7.50 177.50 9.50 3 159 174.75 18.75 172.15 13.75 4 175 173.48 1.82 165.88 9.2 5 190 173.36 16.64 170.44 19.56, 6 205 175.02 29.98 180.22 24.78 7 180 178.02 1.98 192.61 12.61 8 182 178.22 378 186.30 4.30 ‘Sum of absolute deviations: 82.45 98.62 MAD = Hbsviasionel 10.31 12.33 rae Determining the MSE pertrtes ita aa a ately UNLOADED 41 180 175.00 Ed 25.00 2 168 175.50 (-7.5P = 56.25 = 159 474.75 (-15.75) = 248.06 4 175 473.18 (1827 = 3.31 5 190 173.36 I (16.64)? = 276.89 6 205 175.02 (29.987 = 898.80 7 180 178.02 (1.98P = 3.92 8 182 478.22 (3.78% = 14.29 ‘Sum of errors squared = 1,526.52 > Forecast errors) MSE = 4—_______+_ = 1,526 .52/8=190.8 n Peaaen Determining the MAPE eS co acne 1 180 175.00 100(5/180) = 2.78% 2 168 175.50 100(7.5/168) = 4.46% 3 159 174.75 100(15.75/159) = 9.90% 4 175 173.18 100(1.82/175) = 1.05% 5 190 173.36 i 100(16.64/190) = 8.76% 6 205 175.02 100(29.98/205) = 14.62% 7 180 178.02 | 100(1.98/180) = 1.10% 8 182 178.22. 100(3.78/182) = 2.08% ‘Sum of % errors = 44.75% absolute percent error y MAPE = dSolute Percent error _ 44.75% _ 5 coy, as HOA SEN UNIVERSITY Actual Tonnage Quarter Unloaded 1 ONOOARWN 180 168 159 175 190 205 180 182 Rounded Forecast with a= .10 175.00 175.50 174.75 173.18 173.36 175.02 178.02 178.22 Absolute Deviation for a= .10 5.00 7.50 15.75 1.82 16.64 29.98 1.98 3.78 82.45 Rounded Forecast with a= 50 175.00 177.50 172.75 165.88 170.44 180.22 192.61 186.30 Comparison of Forecast Error Absolute Deviation for a= .50 5.00 9.50 13.75 9.12 19.56 24.78 12.61 4.30 98.62 = AY te Comparison of Forecast Error > |deviations| Inded Absolute MAD = ———————_ pcast Deviation n with for = 50 a= .50 For a = .10 175.00 5.00 = 82.45/8 = 10.31 177.50 9.50 172.75 13.75 = 165.88 9.12 Boao 170.44 19.56 = 98.62/8 = 12.33 180.22 24.78 192.61 12.61 SOLE ITO 186.30 4.30 82.45 98.62 MSE = n For a = .10 = 1,526.54/8 = 190.82 For a = .50 = 1,561.91/8 = 195.24 82.45 MAD 10.31 > (forecast errors)? ded cast with 50 175 177.50 172.75 165.88 170.44 180.22 192.61 186.30 “any Comparison of Forecast Error Absolute Deviation for a= .50 5.00 9.50 13.75 9.12 19.56 24.78 12.61 4.30 98.62 12.33 x 00|deviation,|/actual, Absolute MAPE = =" - peviation a= 50 For a = .10 = 44.75/8 = 5.59% For « = .50 = 64.05/8 = 6.76% SStNRATS o a a r 98.62 MAD 10.31 12.33 MSE 190.82 195.24 Pen my OPERATIONS MANAGEMENT = Dees. Comparison of Forecast Error | SEER Rounded Absolute Rounded Absolute Actual Forecast Deviation Forecast Deviation Tonnage with for with for Quarter Unloaded a= .10 a= 10 a= .50 a= 50 HOA SEN UNIVERSITY Fee Trend Adjustment + When atrend is present, exponential smoothing must be modified a Exponential Smoothing with TT ete Rel er Cae AR eT eT ee ae 1 100 F,= 100 (given) 2 200 F,=F, + 0(A, - F,) = 100 + .4(100 — 100) = 100 3 300 F.= Fy + oA, —F,) = 100 + .4(200— 100) = 140 4 400 F,= Fy + o(Ag— Fy) = 140 + .4(300 — 140) = 204 5 500 F,= Fa + o(A,—F,) = 204 + .4(400 — 204) = 282 a voasy EXponential Smoothing with UNIVERSITY CSmREEE Trend Adjustment Forecast Exponentially Exponentially including (FIT,) = smoothed (F,) + smoothed (T,) trend forecast trend F,= ofAy. 1) + (1 - O)(Fp.4 + Ty.) T, = BOF Fy. 4) #4 BT. 4 where F,= exponentially smoothed forecast average T, = exponentially smoothed trend A, = actual demand a = smoothing constant for average (0 < a < 1) B = smoothing constant for trend (0 probability Eas ae distribution 32. o 1 2 8 4 5 6 7 Area payroll (in $ billions) =H AN... Standard Error of the Estimate | SEER where y =y-value of each data point y, = computed value of the dependent variable, from the regression equation n = number of data points =H AN... Standard Error of the Estimate | SEER Computationally, this equation is considerably easier to use We use the standard error to set up prediction intervals around the point estimate a Standard Error of the Estimate gs =.|Y 2d Y~PDY __ [39.5-1.75(15.0)- 2551.5) he n-2 6-2 = V.09375 = .306 (in $ millions) 2 oS e288. Ae f= The standard error Se 20 ° . of the estimate is 32, ' $306,000 in sales OE eS nS Area payroll {in $ billions) et =H eet Session Contents | SEER | 4. Forecasting 4.1. The Nature of Forecasting 4.2. The Strategic, Importance of Forecasting 4.3. How to do the Forcasting Task 4.4. Qualitative Methods porecasong Methods Abb noses Correlation | SEER ¢ How strong is the linear relationship between the variables? ¢ Correlation does not necessarily imply causality! * Coefficient of correlation, f, measures degree of association + Values range from -1 to +1 =H AN 198585, Correlation Coefficient | SEER nSa-SDy Pd¥- Br] %-Oy)| f= AN 198585, Correlation Coefficient | SEER x {0)Perfect positive (a) Perfect nagative o ms correlation carreation Moderate Moderate High 1.0 -08 06 04 02 =H AN 198585, Correlation Coefficient | SEER p E ; a ; 20 1 1 20 40 3.0 3 9 90 90 25 4 16 10.0 6.25 20 2 4 40 40 20 1 1 20 40 35 7 49 245 12.25 y= 150 w= 18 | nde Wsol) \ By= sis qyt= 39.5 (6)(51.5)—(18)(15.0) 18)° |[(16)(39.5) -15 [(6)(0 309-270 39 39 “Yasei2) i872 433° Peaaen =H AN 198585, Correlation | SEER * Coefficient of Determination, r2, measures the percent of change in y predicted by the change in x + Values range from 0 to 1 « Easy to interpret For the Nodel Construction example: r= .901 = 84 = Dees. Multiple-Regression Analysis | SEER + If more than one independent variable is to be used in the model, linear regression can be extended to multiple regression to accommodate several independent variables (Yrar bx +bx, = Computationally, this is quite complex and generally done on the computer = Dees. Multiple-Regression Analysis | SEER In the Nodel example, including interest rates in the model gives the new equation: =1.80+4.30x,-5.0x, An improved correlation coefficient of r = .96 suggests this model does a better job of predicting the change in construction sales Sales = 1.80 + .30(6) - 5.0(.12) = 3.00 Sales = $3,000,000 =H eet Session Contents | SEER | 4. Forecasting 4.1. The Nature of Forecasting 4.2. The Strategic, Importance of Forecasting 4.3. How to do the Forcasting Task 4.4. Qualitative Methods 4.5. Time-Series Forecasting Methods 4.6. Associative Forecasting Methods es Monitoring and Controlling Forecasts UNIVERSITY Tracking Signal « Measures how well the forecast is predicting actual values ¢ Ratio of cumulative forecast errors to mean absolute deviation (MAD) ¢ Good tracking signal has low values + If forecasts are continually high or low, the forecast has a bias error a sasy Monitoring and Controlling Forecasts Tracking _ Cumulative error signal “7 | MAD ¥ (Actual demand in period /-Forecast demand in period /) P }|Actual-Forecast = Dees. Tracking Signal | SEER Signal exceeding limit Tracking signal Upper control limit Lower control limit Time 95 115 100 125 140 Nom yay) Tec DEMAND 100 100 10 10 10 +15 -10 +15 +30 ny -15 10 45 +35 Se Pd FORECAST Ll ear 15 10 15 30 rd le Ud ERROR 15 30 40 55 85 =10/10 = 75 15/7. 10. ono=0 10, 10/10 = -1 11.0 | 45/11 =40.5 14.2 \ +39/14.2= 425 At the end of quarter 6, MAD = Tracking signal = MAD > Forecast errors| 35 n Cumulative error 35 14.2 —=14.2 6 = 2.5 MADs = Dees. Adaptive Smoothing | SEER + It’s possible to use the computer to continually monitor forecast error and adjust the values of the a and B coefficients used in exponential smoothing to continually minimize forecast error ¢ This technique is called adaptive smoothing = Dees. Focus Forecasting | SEER + Developed at American Hardware Supply, based on two principles: 1. Sophisticated forecasting models are not always better than simple ones 2. There is no single technique that should be used for all products or services + Uses historical data to test multiple forecasting models for individual items « Forecasting model with the lowest error used to forecast the next demand as HOA SEN Pro ble m 4. 1

    You might also like