Enhancing Stock Price Forecasting and Trading Strategy through Bidirectional LSTM Integration
Published in 2024 5th International Conference on Machine Learning and Human-Computer Interaction (MLHMI), 2024
Abstract
Accurately predicting data flow is a significant challenge in industrial automation, especially given the diversity of data types. Traditional time series prediction models often struggle to consistently produce effective predictions across varied data sets. To address these limitations, this paper explores time-series prediction models with traditional LSTM and Bidirectional LSTM (Bi-LSTM) architectures. Experimental results indicate that our proposed Bi-LSTM model offers superior prediction accuracy compared to the standalone LSTM algorithm. Furthermore, leveraging our prediction model, we develop an efficient trading strategy algorithm that potentially yields a positive capital gain compared to a holding position.
Authors: Jiayi Liu, Yufeng Yang, Teng Lin, Chuanhui Peng, and Yancong Deng
Published in: 2024 5th International Conference on Machine Learning and Human-Computer Interaction (MLHMI), pp. 22–25
Recommended citation: Jiayi Liu, Yufeng Yang, Teng Lin, Chuanhui Peng, and Yancong Deng. (2024). "Enhancing Stock Price Forecasting and Trading Strategy through Bidirectional LSTM Integration." 2024 5th International Conference on Machine Learning and Human-Computer Interaction (MLHMI), pp. 22–25. DOI: 10.1109/MLHMI63000.2024.00013
Download Paper