MSRN-Informer: Time Series Prediction Model Based on Multi-Scale Residual Network

Time series is a huge quantity of data related to time sequence in real life and its forecast remains challenging. In this study, we propose deep learning model enhance the precision forecast, called MSRN-Informer (Multi-scale Residual Network Improved Informer) model. This can reduce waste significant resources overfitting caused by increasing depth network traditional improvement methods. A multi-scale structure added Informer extract features different scales, residual applied loss. To prove effectiveness presented model, it compared with Informer, Informer+ ARIMA methods on four datasets. The results show that has better prediction ability reduced error. research findings paper be potentially used as reliable reference basis for effective prediction.