Multi-Step Vector Output Prediction of Time Series Using EMA LSTM

This research paper proposes a novel method, Exponential Moving Average Long Short-Term Memory (EMA LSTM), for multi-step vector output prediction of time series data using deep learning. The method combines the LSTM with exponential moving average (EMA) technique to reduce noise in and improve accuracy prediction. compares performance EMA other commonly used learning models, including LSTM, GRU, RNN, CNN, evaluates results statistical tests. dataset this study contains daily stock market prices several years, inputs 60, 90, 120 previous days, predictions next 20 30 days. show that outperforms models terms accuracy, lower RMSE MAPE values. has important implications real-world applications, such as forecasting climate prediction, highlights importance careful preprocessing models.