Comparing Possibility Grey Forecasting with Neural network-based Fuzzy Regression by an Empirical Study

Causality and time series model are the most effective methods used in forecasting practices. Time series models, such as ARIMA, are used by most researchers in stock price prediction. However, in the financial environment, the information on the stock market is vague. To solve this problem, this work presents two forecasting models to help investors make decisions in stock market: one is a new model named possibility grey forecasting model, and the other is the neural network-based fuzzy regression. Moreover, the differences between them and the scenarios for implementing them are also analyzed in this paper to help investors to plan their own investment strategies under various conditions. In the empirical study, we demonstrate that the proposed method and the neural network-based fuzzy regression can be used to effectively find the stock index in Taiwan.