Revolutionizing Wealth Management with Temporal Attention Mechanism Based Convolutional Network for Algorithmic Stock Trading Strategies

Stock trading strategies are systematic plans utilized by investors or traders to buy and sell stocks for profit, relying on technical and fundamental analysis or a combination of both. However, algorithmic stock trading operates within big data and unpredictable financial markets, making accurate stock price forecasting challenging due to factors such as investor behavior and economic indicators. These factors impact stock trading strategies by influencing risk management and return maximization. Robo-advisors are transforming wealth management by combining big data and Artificial Intelligence (AI) within algorithmic trading strategies. They leverage AI technologies to analyze vast amounts of data, identify market patterns, and optimize investment portfolios. To overcome these challenges, the proposed Temporal Attention Mechanism – Temporal Convolutional Network (TAM-TCN) is utilized for accurate stock forecasting. The proposed TAM-TCN improves stock price prediction through a temporal attention mechanism that captures complex feature dependencies alongside TCN. Min-max normalization is performed during preprocessing, and Moving Average (MA) is employed in feature selection to enhance the precision of algorithmic stock price prediction. The proposed method achieves a cumulative return (CR) level of 591.25% compared to Deep Neural Networks on the IXIC dataset.