Forecasting demand and inventory management using Bayesian time series

Abstract

Discusses the development and evaluation of a forecasting model for inventory management in an advanced technology batch production environment. Traditional forecasting and inventory management do not adequately address issues relating to a short life cycle and to non‐seasonal products with a relatively long lead time. Limited historical data (fewer than 100 observations) is also a problem in predicting short‐term dynamic or unstable time series. A Bayesian dynamic linear time series model is proposed as an alternative technique for forecasting demand in a dynamically changing environment. Provides details of the important characteristics and development process of the forecasting model. A case study is then presented to illustrate the application of the model based on data from a multinational company in Singapore. It also compares the Bayesian dynamic linear time series model with a classical forecasting model (auto‐regressive integrated moving average (ARIMA) model).

Citation

Spedding, T.A. and Chan, K.K. (2000), "Forecasting demand and inventory management using Bayesian time series", Integrated Manufacturing Systems, Vol. 11 No. 5, pp. 331-339. https://doi.org/10.1108/09576060010335609