Evaluation of subjective forecasting in electrical utilities

This thesis is an investigation into the effectiveness of some mathematical forecasting methods as compared with subjective forecasting methods. The author points out that one mathematical forecasting method, triple smoothing, projects as well as the subjective forecast made by an electrical utility when forecasting long-range annual power demands. He further hypothesizes that by using the triple smoothing method coupled with knowledge of some major customers that an electrical utility can develop better forecasts than those made subjectively.

The thesis also includes a discussion of some basic problems encountered in forecasting and in evaluating forecasts.