Use of Machine Intelligence in Over-the-Counter Products in Stock Market

In this study, machine intelligence was applied to predict stock prices of over-the-counter products and to compare the performance of traditional learning models. The research aimed to assess the effectiveness of these stocks and examine various influencing factors such as profitability, trading volume, and market sentiment. Model inferences were evaluated by combining historical stock price data, sentiment analysis, and profitability metrics. The results show that long short-term memory models outperformed traditional models in accuracy, precision, recall, and profitability, achieving a 9.5% average monthly return and a Sharpe ratio of 2.1. Moreover, the introduction of machine intelligence models increased trading volume, demonstrating their impact on trading activity. Both sentiment analysis and stock price movements exhibited a strong correlation. Overall, the findings indicate that long short-term memory-based deep learning models outperform traditional methods in prediction and profitability within the over-the-counter stock markets, with significant implications for trading strategies.