In today's world, where financial and psychological data are abundantly available (such as social media data, economic and analytical news, and search trends), the use of machine learning methods as a powerful tool for analyzing complex and large data sets seems essential. Machine learning not only has the ability to identify hidden patterns in data, but also can model nonlinear and multidimensional relationships with high accuracy. In financial markets, investor behavior plays a significant role in stock price fluctuations and returns. This study investigates the impact of investor psychology on stock price behavior in the Tehran Stock Exchange using machine learning methods. For modeling purposes, the market sentiment index was extracted based on historical data from social networks, trading volume, and changes in the total index. Then, using machine learning algorithms such as random forest, support vector machine, and artificial neural networks, the impact of sentiment on stock volatility and returns was analyzed. The results show that investor sentiment has a significant impact on stock price behavior and the artificial neural network algorithm is able to predict return changes with higher accuracy than other methods. The findings of this study can help investors make better decisions and improve financial policies.