A Hybrid Model Using Coordinate Attention Mechanisms for Stock Price Prediction

Predicting stock prices is a complex and challenging task in financial markets due to the intricate and non-linear nature of stock price movements. This study explores the application of Coordinate Attention mechanisms within advanced Temporal Convolutional Network (TCN) frameworks for stock price prediction based solely on historical stock price data. We present a hybrid model that combines the advantages of modern TCNs with Coordinate Attention mechanisms to improve the precision and robustness of stock prediction models. Extensive experiments are conducted on the well-known TSLA stock dataset. The results show varying levels of alignment with actual stock prices, highlighting the strengths and weaknesses of each model.