Scholarlink

The stock market is an energetic and agile research area, and predicting its complexion is an epic fundamental nowadays. The stock market is chaotic, complex, volatile, and dynamic in nature. Indeed, its prediction is a trivial task in time series analysis. In this article, we have taken several Neural Network (NN) models like Recurrent Neural Network (RNN), Stacked RNN, Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), Bidirectional GRU, and Bidirectional LSTM (Bi-LSTM) for forecasting stock amounts. The historical data for an indexed stock in the National Stock Exchange (NSE, India), TATA Motors Ltd., has been taken into consideration to train the mentioned models. Their performance was evaluated for predicting future stock prices. After observing the results, it is concluded that the Bidirectional GRU model accomplished better than the other models in the performance measuring metric, root mean squared error (RMSE).

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

Performance Analysis of Machine Learning Models in Short-Term Stock Predictions on the Indian Market

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Sohini Mukherjee,
  • Soham Dutta,
  • Arpan Adhikary,
  • Asit Kumar Nayek

摘要

The stock market is an energetic and agile research area, and predicting its complexion is an epic fundamental nowadays. The stock market is chaotic, complex, volatile, and dynamic in nature. Indeed, its prediction is a trivial task in time series analysis. In this article, we have taken several Neural Network (NN) models like Recurrent Neural Network (RNN), Stacked RNN, Gated Recurrent Unit (GRU), Long Short-Term Memory (LSTM), Bidirectional GRU, and Bidirectional LSTM (Bi-LSTM) for forecasting stock amounts. The historical data for an indexed stock in the National Stock Exchange (NSE, India), TATA Motors Ltd., has been taken into consideration to train the mentioned models. Their performance was evaluated for predicting future stock prices. After observing the results, it is concluded that the Bidirectional GRU model accomplished better than the other models in the performance measuring metric, root mean squared error (RMSE).