e INVENTORY ANALYTICS

In this section, we will focus on four simple forecasting methods and on their underlying stochastic models.27 27 The four methods are: the Moving Average method; the Naïve method; the Drift method; and the Seasonal Stationary demand & the Moving Average method Naïve method. [...] In essence, the mean µ of all future random variables is assumed to be equal to the average of all historical realisations (Average method), or to the average of the past w realisations (Moving Average method). [...] 44 Forecasts for the last 40 periods by using the Moving Average method with a window of size w = 32; the underpinning stochastic process is a a Gaussian process with mean µ = 20 and standard deviation σ = 5. [...] .; this variant of the Naïve method allows the forecasts to increase or decrease over time, where the amount of change over time (the drift) is set to be the average change seen in the historical data. [...] A forecast method that minimises the MAE will lead to forecasts of the median, while minimising the RMSE will lead to forecasts of the mean.