Statistical Methods for Forecasting

Statistical Methods for Forecasting is a comprehensive, readable treatment of statistical models and methods used to produce short-term forecasts. Bridging the gap between introductory, descriptive approaches and highly advanced theoretical treatises, it provides a practical intermediate-level discussion of a venery of forecasting tools, and explains how they relate to one another both in theory and practice. While the emphasis is on the familiar regression models, and exponential smoothing and parametric time series models for nonseasonal and seasonal data, the text also treats a number of special topics such as transfer function analysis, Kalman filtering, state space models, Bayesian forecasting, seasonal adjustment and forecast evaluation. a unique feature of the presentation is the interrelation of forecasts from exponential smoothing and forecasts from ARIMA (autoregressive integrated moving average) time series models. This discussion shows which ARIMA models imply the various exponential smoothing forecast procedures as special cases. The text also adopts a model-based approach to forecasting, one which uses available data to construct appropriate models. Statistical Methods for Forecasting serves as an outstanding textbook for graduate and advanced undergraduate courses in forecasting for students of statistics, mathematics, business, engineering, and the social sciences, as well as a basic working reference for professional forecasters in business, industry, and government. It includes a large number of examples and exercises (using real data) and provides numerous time series, autocorrelation and partial autocorrelation plots as illustrations.Abraham, Bovas is the author of 'Statistical Methods for Forecasting', published 1983 under ISBN 9780471867647 and ISBN 0471867640.