An Improved Heuristic-Based Fuzzy Time Series Forecasting Model Using Genetic Algorithm

Fuzzy time series is being used for forecasting since last two decades and a lot of work has been done by different researchers to get better forecasting models and higher forecasting accuracy. In this paper a heuristic trend predictor is proposed based on fuzzy time series forecasting model. The model will uses genetic algorithm for adjusting interval length to get improved results. The proposed method will be applied on car road accidents causalities data of Belgium and hopefully will get better results than many other previous methods.