An ARIMA- LSTM Hybrid Model for Stock Market Prediction Using Live Data
Kulshreshtha Sakshi, A. Vithya Vijayalakshmi
Vellore Institute of Technology University
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
The stock market is a highly volatile industry with ever changing bull (rise) and bear (fall) trends.This paper proposes a new hybrid model using Long Short-Term Memory (LSTM), a Recurrent Neural Network (RNN) technique and Auto Regressive Integrated Moving Average (ARIMA), a time series forecasting technique to capture the live stock market data of S&P 500 using preexisting Application Programming Interface (API).Rise and fall in stock values in the previous years is analyzed.A novel LSTM-ARIMA hybrid is designed for capturing the linear and non-linear portions of the time series.The Prophet forecasting library by Facebook has also been used that requires less preprocessing.Finally, both the approaches are compared and the one with better performance is accepted for the final stock market prediction system.In this case, Prophet has a high Root Mean Square Error (RMSE) of 27.59 and Mean Square Error (MSE) of 761.33 whereas the ARIMA-LSTM hybrid gives an MSE of 3.03 and RMSE of 1.74 along with a 99% fit of the model.Hence the hybrid performs much better than Prophet and is accepted as the final algorithm for implementation.
逐年被引趋势
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
计算机 / AIStock Market Forecasting Methods
Forecasting Techniques and Applications · Time Series Analysis and Forecasting
参考文献 12
此处列出前 3 条
引用本文 56
按被引量排序,此处列出前 3 条