Scholarlink

The fluctuation law of the stock market, a highly complex nonlinear movement system, is influenced by numerous causes, including predicting the stock price index is hence a highly challenging task. Numerous instances demonstrate how well-suited neural network algorithms are for making these kinds of time series predictions, and they frequently produce good outcomes. Based on the existing models, we proposed a Regularized GRU Regularized LSTM neural network model in this study and used it to the short-term stock closing price projection. The trial’s results show that, for stock time series prediction, our recommended model performs better than the GRU and LSTM models of networks currently in use.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep Learning for Stock Prediction: Comparative Analysis of Recurrent Neural Network Architectures and Regularization Techniques

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Vara Prasad Lakshmi,
  • Venkata Rao,
  • Kishore Babu,
  • Vamsi Krishna Lakshmi,
  • Sharada Mani,
  • Madhu Raghavulu

摘要

The fluctuation law of the stock market, a highly complex nonlinear movement system, is influenced by numerous causes, including predicting the stock price index is hence a highly challenging task. Numerous instances demonstrate how well-suited neural network algorithms are for making these kinds of time series predictions, and they frequently produce good outcomes. Based on the existing models, we proposed a Regularized GRU Regularized LSTM neural network model in this study and used it to the short-term stock closing price projection. The trial’s results show that, for stock time series prediction, our recommended model performs better than the GRU and LSTM models of networks currently in use.