GATE: A guided approach for time series ensemble forecasting

Md. Rasel Sarkar, Sreenatha G. Anavatti, Tanmoy Dam, Md Meftahul Ferdaus, Murat Tahtali, Savitha Ramasamy, Mahardhika Pratama

Abstract: Highlights•RNN, LSTM, and Conv-LSTM blend decodes time series features and prevents overfitting.•The guided network identifies optimal models during training.•Error metrics enhance feature extraction from diverse samples.•Outperforms contemporaries with superior accuracy on 4 real-world datasets.