MODELING INVESTOR BEHAVIOR AND INVESTMENT DECISIONS IN INDONESIA’S TECHNOLOGY SECTOR: A HYBRID FUZZY DEMATEL- MACHINE LEARNING FRAMEWORK

Abstract

This study aims to examine investor behavior and investment decision-making in Indonesia’s technology sector by integrating Fuzzy Decision-Making Trial and Evaluation Laboratory (Fuzzy DEMATEL) and Machine Learning within a hybrid analytical framework. A mixed-data approach employed secondary data from 22 technology companies listed on the Indonesia Stock Exchange during 2020-2024 and primary data from seven finance and investment experts. The analysis covered investor sentiment, market volatility, herding behavior, overconfidence, risk perception, stock return, financial leverage, and investment decisions. Fuzzy DEMATEL identified causal relationships, while Random Forest and XGBoost assessed predictive performance and feature importance. The results show that investor sentiment is the primary causal driver and strongest predictor of investment decisions, whereas market volatility acts as the central link between behavioral and financial factors. XGBoost achieved superior predictive accuracy (R² = 0.94). The integrated framework advances behavioral finance research by combining causal analysis with predictive modeling to better explain investment decision-making in emerging technology markets.