A 3D Membership Function-Based Type-2 Fuzzy Brain Emotional Learning Predictor for Forecasting Taiwan Stock Price

This study proposes a new efficient predictor called a three-dimensional (3D) membership function-based Type-2 Fuzzy Brain-Emotional Learning Predictor (T2FBELP) to model and predict Taiwan stock prices. This study has primarily contributed to improve the learning ability and flexibility of type-2 fuzzy system by changing the structure in the membership function space. The use of a 3D Gaussian membership function in the fuzzy network allows for fewer rules while maintaining accurate predictions, especially when dealing with nonlinear time series such as stock prices. Additionally, learning-rate parameters are optimized using the Modified Social Ski-Driver (MSSD) algorithm to further impoving the predict ability. The T2FBELP is then used to predict the stock prices for including the Taiwan Semiconductor Manufacturing Company (TSMC), the Honhai Precision Industry and the Formosa Plastics Corporation. These efforts show the effectiveness of T2FBELP in predicting the Taiwan stock market.