To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Chapter 9 Forecasting
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Patterns of Demand Quantity Time (a) Horizontal: Data cluster about a horizontal line. Figure 9.1
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Patterns of Demand Quantity Time (b) Trend: Data consistently increase or decrease. Figure 9.1
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Patterns of Demand Quantity |||||||||||| JFMAMJJASOND Months Year 1 Year 2 (c) Seasonal: Data consistently show peaks and valleys. Figure 9.1
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Patterns of Demand Quantity |||||| Years (c) Cyclical: Data reveal gradual increases and decreases over extended periods. Figure 9.1
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Demand Forecast Applications Time Horizon Medium TermLong Term Short Term (3 months -(more than Application(0-3 months) 2 years) 2 years) Forecast quantityIndividualTotal salesTotal sales products orGroups or families servicesof products or services Decision areaInventoryStaff planningFacility location managementProductionCapacity Final assemblyplanningplanning schedulingMaster productionProcess Work-forceschedulingmanagement schedulingPurchasing Master productionDistribution scheduling ForecastingTime seriesCausalCausal techniqueCausalJudgmentJudgment Judgment Table 9.1
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Causal Methods Linear Regression Dependent variable Independent variable X Y Actual value of Y Estimate of Y from regression equation Value of X used to estimate Y Deviation, or error { Regression equation: Y = a + bX Figure 9.2
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Time Series Methods Simple Moving Averages Figure 9.4 Week 450 — 430 — 410 — 390 — 370 — |||||| Actual patient arrivals Patient arrivals
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Time Series Methods Simple Moving Averages Example 9.2 Actual patient arrivals Actual patient arrivals 450 — 430 — 410 — 390 — 370 — Week |||||| Patient WeekArrivals Patient arrivals
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Time Series Methods Simple Moving Averages 450 — 430 — 410 — 390 — 370 — Week |||||| Actual patient arrivals 3-week MA forecast Patient arrivals
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Time Series Methods Exponential Smoothing Example — 430 — 410 — 390 — 370 — Week |||||| F 4 = Exponential Smoothing = 0.10 F 3 = ( )/2 D 3 = 411 F t +1 = F t + (D t - F t ) Patient arrivals
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Time Series Methods Exponential Smoothing Example 9.3 Week 450 — 430 — 410 — 390 — 370 — |||||| F 4 = D 4 = 415 Exponential Smoothing = 0.10 F 4 = F 5 = F t +1 = F t + (D t - F t ) Patient arrivals
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Time Series Methods Exponential Smoothing 450 — 430 — 410 — 390 — 370 — Patient arrivals Week |||||| Example 9.3 Exponential smoothing = 0.10
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Measures of Forecast Error E t = D t - F t Choosing a Method Forecast Error Example 9.4
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Choosing a Method Forecast Error Measures of Forecast Error E t = D t - F t |E t | n Et2nEt2n CFE = E t = MSE = MAD = MAPE = [ |E t | (100) ] / D t n (E t - E ) 2 n - 1 Example 9.4
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Absolute Error AbsolutePercent Month,Demand,Forecast,Error,Squared,Error,Error, tD t F t E t E t 2 |E t |(|E t |/D t )(100) % Total % Choosing a Method Forecast Error Example 9.4
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Choosing a Method Forecast Error Absolute Error AbsolutePercent Month,Demand,Forecast,Error,Squared,Error,Error, tD t F t E t E t 2 |E t |(|E t |/D t )(100) % Total % MSE = = CFE = - 15 Measures of Error MAD = = MAPE = = 10.2% 81.3% 8 E = = = 27.4 Example 9.4
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Choosing a Method Tracking Signals Table 9.2 Percentage of the Area of the Normal Probability Distribution Within the Control Limits of the Tracking Signal Control Limit SpreadEquivalentPercentage of Area (number of MAD)Number of 2 Within Control Limits ± 1.0± ± 1.5± ± 2.0± ± 2.5± ± 3.0± ± 3.5± ± 4.0±
To Accompany Ritzman & Krajewski, Foundations of Operations Management © 2003 Prentice-Hall, Inc. All rights reserved. Choosing a Method Tracking Signals Tracking signal = CFE MAD +2.0 — +1.5 — +1.0 — +0.5 — 0 — — — — ||||| Observation number Tracking signal Control limit Out of control Figure 9.5