Scholarlink

Using machine learning and deep learning approaches, this research study investigates the process of stock price prediction. A hybrid model comprising of ARIMA, i.e., Auto Regressive Integrated Moving Average, a time series forecasting technique and a Recurrent Neural Network (RNN) technique called Long Short Term Memory (LSTM), was developed for the purpose of forecasting the stock prices, as hybrid models combine the strengths of multiple algorithms, mitigating their individual weaknesses. In order to obtain more accurate predictions, hyperparameter tuning was also done using Optuna, a software framework for autonomous hyperparameter optimization. Within the field of machine learning, the prediction of stock prices is a significant and quickly growing area of research, which creates new opportunities for academics to explore further.

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

Study on Stock Price Forecasting Using a Hybrid ARIMA-LSTM Model

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Soumodip Kumar Paul,
  • Kuldip Katiyar

摘要

Using machine learning and deep learning approaches, this research study investigates the process of stock price prediction. A hybrid model comprising of ARIMA, i.e., Auto Regressive Integrated Moving Average, a time series forecasting technique and a Recurrent Neural Network (RNN) technique called Long Short Term Memory (LSTM), was developed for the purpose of forecasting the stock prices, as hybrid models combine the strengths of multiple algorithms, mitigating their individual weaknesses. In order to obtain more accurate predictions, hyperparameter tuning was also done using Optuna, a software framework for autonomous hyperparameter optimization. Within the field of machine learning, the prediction of stock prices is a significant and quickly growing area of research, which creates new opportunities for academics to explore further.