Past and Future of Demand Forecasting Models

This chapter sheds some light on how the demand forecast impacts business decisions at the strategic, tactical, and operational levels. We classify demand forecasting models according to demand dependency as independent-demand models (i.e., time-series models) and dependent-demand models (i.e., causal models). For the independent models, we discuss the most well-known forecasting techniques known as time-series forecasting techniques. We classify the time-series forecasting techniques based on existing patterns, such as level, trend, seasonality, and lumpiness, in the time series. For the dependent models, we discuss regression and simulation-based models employed for parameter estimation of the dependent-demand models. Furthermore, we discuss scale-dependent and scale-independent forecast accuracy metrics as well as the accuracy of the historical data. Finally, we address how emerging technologies such as Big Data and artificial intelligence in the form of machine learning and deep learning algorithms can play a substantial role in enhancing the demand forecasting process for both classes of the demand models.