Integrating Fuzzy Logic with Machine Learning for Intelligent Stock Trading Systems

Abstract

This study presents a decision support system for stock market investors by integrating fundamental analysis with fuzzy logic. The system utilizes four key indicators—Risk, Profitability Ratios, Earnings per Share (EPS), and the Price-to-Earnings (P/E) Ratio—as inputs to a fuzzy inference model. Through predefined fuzzy rules and membership functions, the model generates trading recommendations categorized as Bullish, Neutral, or Bearish. An experimental analysis conducted in MATLAB on stock data from the top 10 NASDAQ-listed companies demonstrates the system’s effectiveness in predicting market movements and improving investment decisions when complemented with market insights and investor perspectives. The results indicate that the proposed model successfully predicts daily trading signals, ranking Google in the top position, while NVIDIA received the lowest rank, reflecting relatively weaker performance among the companies analyzed. Overall, the system offers valuable support for both individual investors and mutual fund managers in making informed, data-driven decisions.