Fuzzy logic based time series prediction algorithm using nearest neighborhood clustering

Abstract EN

This paper presents a fuzzy logic based prediction algorithm using nearest neighborhood clustering for the prediction of time series.

The proposed algorithm has been developed by using nearest neighborhood clustering scheme and optimal fuzzy system.

The presented algorithm is tested by the prediction of Mackey-Glass time series, time series obtained from Logistic Map equation, and an oil refinery data series.

To the best of our knowledge, very few studies have attempted to predict Logistic Map.

The simulation results show that the proposed algorithm is able to predict linear, nonlinear, and even chaotic time series to a reasonable accuracy.

It performs well, even when the mathematical model of time series does not exist.