The Research on End-to-end Stock Recommendation Algorithm Based on Time-frequency Consistency

In the financial market, the volatility and complexity of stock prices make accurately predicting stock trends a highly challenging task. Traditional stock prediction methods often rely on either time-domain or frequency-domain information alone, which fails to fully capture the multi-scale dynamic characteristics of stock prices, leading to insufficient prediction accuracy. To address the shortcomings of existing stock recommendation algorithms, this paper proposes an end-to-end stock recommendation algorithm based on time-frequency consistency. Firstly, we introduce a time-frequency consistency analysis method, which can simultaneously extract both time-domain and frequency-domain features of stock prices, thus providing a more comprehensive characterization of stock trend changes. Secondly, by integrating prompt learning strategies, the model is guided by pre-designed prompts to identify the lowest-risk buying points within specific time frames, optimizing the stock recommendation decision-making process. Finally, the end-to-end model training ensures seamless integration and automation throughout the entire prediction process, achieving a complete workflow from data input to stock recommendation. Experimental results demonstrate that this method outperforms traditional approaches in terms of prediction accuracy and risk control, offering more reliable decision support for investors.