IEOR 165 Discussion 10 April 24, 2015 Question 1. Assume that the demand is: Year 1 1 2 3 4 Demand 10 20 26 17 Year 2 1 2 3 4 Demand 12 23 30 22 a. Using the data, determine initial values of the intercept, slope and seasonal factors for multiplicative Winter’s method. b. Assume that the observed demand for the first quarter of year 3 was 16. Using α = 0.2, β = 0.1 and γ = 0.1, update the estimates of the series, the slope and the seasonal factors. 1 Solution 2. a. Seasonal factor initialization Dt Vi − [(N + 1)/2 − j]G0 where i = 1 for the first season and i = 2 for the second season and j is the period of the season. Initialization procedure: ct = V1 1 = N V2 = G0 = S0 = c−7 = c−6 = c−5 = c−4 = c−3 = c−2 = c−1 = c0 = c−3 c−2 c−1 c0 sum: 1 N −N ∑ 1 Dj = (10 + 20 + 26 + 17) = 18.25 4 j=−2N +1 0 ∑ 1 Dj = (12 + 23 + 30 + 22) = 21.75 4 j=−N +1 V2 − V1 21.75 − 18.25 = = 0.875 N 4 3 N −1 = 21.75 + (0.875) = 23.06 V2 + G0 2 2 10 = 0.5904 18.25 − [5/2 − 1]0.875 20 = 1.123 18.25 − [5/2 − 2]0.875 26 = 1.391 18.25 − [5/2 − 3]0.875 17 = 0.869 18.25 − [5/2 − 4]0.875 12 = 0.5872 21.75 − [5/2 − 1]0.875 23 = 1.079 21.75 − [5/2 − 2]0.875 30 = 1.352 21.75 − [5/2 − 3]0.875 22 = 0.9539 21.75 − [5/2 − 4]0.875 avg seasonal factors (c−7 + c−3 )/2 = 0.5888 (c−6 + c−2 )/2 = 1.1010 (c−5 + c−1 )/2 = 1.372 (c−4 + c0 )/2 = 0.9115 3.9733 2 normalized seasonal factors 0.59 1.11 1.38 0.92 4 Suppose we wish to forecast the following year’s demand at t = 0. The forecasting equation is: Ft,t+τ = (St + τ Gt )ct+τ −N which results in F0,1 F0,2 F0,3 F0,4 = = = = (S0 + 1G0 )c−3 = (23.06 + 0.875)0.59 = 14.12 (S0 + 2G0 )c−2 = [23.06 + (2)(0.875)]1.11 = 27.54 (S0 + 3G0 )c−1 = [23.06 + (3)(0.875)]1.38 = 35.44 (S0 + 4G0 )c0 = [23.06 + (4)(0.875)]0.92 = 24.38 b. α = 0.2 β = 0.1 γ = 0.1 D1 = 18 S1 = α(D1 /c−3 ) + (1 − α)(S0 + G0 ) = 0.2(16/0.59) + 0.8(23.06 + 0.875) = 24.57 G1 = β(S1 − S0 ) + (1 − β)G0 = 0.1(24.57 − 23.06) + 0.9(0.875) = 0.9385 c1 = γ(D1 /S1 ) + (1 − γ)c−3 = 0.1(16/24.57) + 0.9(0.59) = 0.5961 At this point it is recommended to renorm c−2 , c−1 , c0 and the new value of c1 to add to 4. c1 c−2 c−1 c0 sum: avg seasonal factors 0.5961 1.11 1.38 0.92 4.0061 normalized seasonal factors 0.59 1.11 1.38 0.92 4 Forecasts for 2nd , 3rd and 4th quarters of 1993 F1,2 = [S1 + G1 ]c−2 = (24.57 + 0.9385)1.11 = 28.3144 F1,3 = [S1 + 2G1 ]c−1 = [24.57 + 2(0.9385)]1.38 = 36.4969 3