Abstract
In this paper, we report the development of a method for forecasting enrollments at the University of Alabama using the harmony search algorithm and fuzzy time series. In recent years, researchers have proposed many methods of forecasting, including forecasting enrollments at the University of Alabama or financial markets such as FOREX. In recent research efforts, it was determined that the lengths of the intervals in the universe of discourse and forecasting rules influence the accuracy of forecasting. Some forecasting methods use meta-heuristic algorithms (such as the genetic algorithm) to determine the optimum length of the intervals in the universe of discourse. In this paper, we propose a new, meta-heuristic algorithm (i.e., the harmony search algorithm) that is based on a high-order, fuzzy time series and overlapping intervals. In previous methods, it was assumed that the universe of discourse is partitioned into separate intervals that do not overlap. Here, it is shown that overlapping intervals in the universe of discourse can influence the accuracy of forecasted values.