A Multi-granularity & Fuzzy CBR Based Method for Price Forecast

Abstract

The ability to scientifically and accurately forecast product price fluctuations plays a crucial role in the development of the contemporary market economy and consequently results in vigorous research in the area. This paper presents a new forecast method using multi-granularity of time and case-based reasoning (CBR) with fuzzy matching of cases featuring certain fluctuations. The method has been applied in aquaculture with aquatic product prices as test data. Results of experiments have proved the method highly feasible and accurate in aquatic product price forecasts.