High Order Time Series Forecasting using Fuzzy Discretization

2. Previous Works

Fuzzy set theory introduced by Lotfi A. Zadeh (1965, 1975), is based on the notion of partial containment. Fuzzy characteristic function is related to vagueness. In fuzzy logic degree of membership of a variable has a truth value that ranges between 0 and 1. The advantage of fuzzy forecasting is that it can handle problems using numerical as well as linguistic information such as low, moderate, high, very high, etc.

Among the earlier important research works on prediction were generated by Sugeno and Tanaka (1991)Wang and Mendel (1992). Song and Chissom (1993a, 1993b, 1994) developed the time variant and time invariant models for fuzzy time series forecasting. Further, many researchers, Chen (1996), Sullivan and Woodall (1994), Kim and Lee (1999), Chen and Hwang (2000), Huarng (2001), and Tsai and Wu (2001), worked on the development of various models of fuzzy time series forecasting and its implementations.