AI-Driven Intraday Trading: Applying Machine Learning and Market Activity for Enhanced Decision Support in Financial Markets

In response to the unpredictable fluctuations in the global economic landscape, and with the aim of mitigating overnight risks, the stock market has seen a substantial increase in the number of day traders. However, day traders often fall prey to emotional influences, resulting in abrupt and irrational shifts in market prices. Consequently, when supporting investors engaged in day trading, it becomes imperative to furnish them with decision-making assistance to reduce trading risks. To address this challenge, the present study leverages a neural network architecture paired with a daily market activity structure. This combination allows us to delve into the dynamic behaviors governing day trading in Taiwan’s weighted index futures. By uncovering the underlying knowledge rules of the futures market, we can establish a model for predicting day-trading directions and employ effective trading strategies. The outcomes of this research indicate that the accuracy attained through this research methodology surpasses that of the random walk theory used by the control group. The discernible divergence in results confirms that lower-risk entry points within the intraday market can be identified through this approach. Recognizing these low-risk entry points holds the potential to aid investors in managing market risks and enhancing their opportunities for profit.