Time Series hybrid Prediction Model Based on Deep Learning ARIMA-LSTM
Dongmei Wu · 2024
This study investigates a prediction model for time series data based upon deep learning techniques, addressing the challenges of efficient processing and precise prediction of time series data in practical applications. It presents a hybrid deep learning model that integrates the Autoregressive Integrated Moving Average Model (ARIMA) with the Long Short-Term Memory Model (LSTM) to enhance the predictive capability for time series data. Through comprehensive experiments, this study evaluates the effectiveness of the proposed ARIMA-LSTM hybrid model for time series prediction. The results demonstrate that, compared to traditional forecasting methods, this deep learning hybrid model effectively addresses the limitations of Recurrent Neural Networks (RNNs) in processing long sequences. Furthermore, it enhances the ability to capture long-term dependencies in time series data, thereby significantly improving prediction accuracy. The results presented in this study offers novel insights and methodologies for accurate time series prediction, contributing significantly to the practical application of deep learning technologies.