John E. Boylan - Intermittent Demand Forecasting

5 Chapter 7Table 7.1 Updating of mean and variance using SES.Table 7.2 Updating of variance over protection interval: scaled and direct.Table 7.3 Distributions of demand over gamma distributed lead times.

6 Chapter 8Table 8.1 Safety factors for CSL targets, normal demand.Table 8.2 Safety factors for fill rate (FR) targets, normal demand.Table 8.3 Asymmetric effect of under‐ and over‐forecasting.Table 8.4 Adjusted safety factors for cycle service levels.Table 8.5 Cycle service level for Poisson demand (( R + L ) ).Table 8.6 Fill rate for Poisson distributed demand.Table 8.7 Cycle service levels for stuttering Poisson distributed demand.Table 8.8 Weighted cumulative probabilities.Table 8.9 Adjusted safety factors for fill rates.Table 8.10 component calculations for Poisson distributed demand.Table 8.11 calculations for Poisson distributed demand.Table 8.12 Fill rate calculations for Poisson demand.Table 8.13 component calculations for stuttering Poisson demand.

7 Chapter 9Table 9.1 Mean error, mean square error, mean absolute error, and mean absolu...Table 9.2 Forecast value added (FVA) example.Table 9.3 MAPEFF and sMAPE for intermittent demand.Table 9.4 MAE : Mean ratios for multiple series.Table 9.5 Mean absolute error for zero forecasts.Table 9.6 Mean error (ME) and mean absolute error (MAE).Table 9.7 Scaled mean error for multiple series.

8 Chapter 10Table 10.1 Reported usage of forecast methods in practice.Table 10.2 Judgemental adjustments: effect on cycle service levels.Table 10.3 Cumulative forecast error (CFE).Table 10.4 Mean square error (frequent zeroes).

9 Chapter 13Table 13.1 Cumulative frequency percentages.Table 13.2 Three‐month overlapping blocks (OB) and non‐overlapping blocks (NO...Table 13.3 Resampling from previous observations.Table 13.4 VZ resampling method ( ).Table 13.5 Most recent 10 observations from Table 13.2.Table 13.6 Conditional probabilities of demand occurrence.Table 13.7 Simple bootstrapping with Markov chain extension..Table 13.8 Theta function calculation ( , ), overlap of one period.

10 Chapter 14Table 14.1 INAR(1) process example.Table 14.2 INMA(1) process example.Table 14.3 Four simplest INARMA models.Table 14.4 Empirical evidence on model identification.Table 14.5 Conditional probabilities of demand at time ( ) given demand at t...Table 14.6 Cumulative conditional probabilities at time ( ) given demand at ...Table 14.7 Cumulative probabilities of demand over two periods ( ), given dem...

11 Chapter 15Table 15.1 Software implementation.

1 Chapter 1 Figure 1.1 Intermittent and lumpy demand.

2 Chapter 2 Figure 2.1 Bill of materials (BoM) example. Figure 2.2 Periodic review and continuous review systems. Figure 2.3 Continuous review and policies for unit sized transactions. Figure 2.4 Periodic review policy.

3 Chapter 3 Figure 3.1 Comparison of CSL and . Figure 3.2 Exchange curve. Figure 3.3 RightStock Inventory Strategist.

4 Chapter 4 Figure 4.1 Monthly demand time series for an automotive SKU. Figure 4.2 Demand frequencies for an automotive SKU. Figure 4.3 Demand relative frequencies with triangle superimposed. Figure 4.4 Actual relative frequencies and triangular probabilities.Figure 4.5 Poisson distribution for varying mean ( ) values.Figure 4.6 Poisson probabilities and actual relative frequencies.Figure 4.7 Variance and mean of weekly order frequencies.

5 Chapter 5Figure 5.1 Geometric distribution ( ).Figure 5.2 Standard deviation and mean of demand sizes. Source: Johnston et ...Figure 5.3 Frequency distribution of order sizes. Source: Johnston et al. 20...Figure 5.4 Logarithmic distribution ( = 0.33, 0.66, 0.99).Figure 5.5 Exponential distributions. (a) Probability density; (b) Cumulativ...Figure 5.6 Erlang distributions.Figure 5.7 Normal distribution (poor approximation).Figure 5.8 Normal distribution (better approximation).

6 Chapter 6Figure 6.1 Weights of previous observations. (a) . (b) .Figure 6.2 SES response to a step‐change. (a) . (b) .Figure 6.3 SES bias for issue points only ( ).Figure 6.4 Forecast initialisation and optimisation.Figure 6.5 ADIDA forecasting framework.Figure 6.6 Comparison of model forms.

7 Chapter 8Figure 8.1 Standard normal distribution.Figure 8.2 Normally distributed demand and OUT levels.

8 Chapter 9Figure 9.1 Errors and absolute (‘Abs’) errors.Figure 9.2 Non‐uniform distributions of randomised PITs.Figure 9.3 Exchange curves.

9 Chapter 10Figure 10.1 Cumulative demands and forecasts.Figure 10.2 Squared error decomposition.Figure 10.3 (Extended) squared error decomposition.