Stacked Ensemble Model for Tropical Cyclone Path Prediction

Tropical cyclones are intense circular storms that cause significant economic and life losses in the coastal areas of equatorial region. Various statistical models were proposed to forecast tropical cyclone’s potential path. This study has a stacked ensemble-based method increase temporal data’s Cyclone path prediction effectiveness. The can be divided into two phases; first phase, Long Short-Term Memory Networks(LSTM) Gated Recurrent Unit(GRU) optimized with layers investigated best possible for Stacked LSTM GRU. In second k-fold cross-validation is used construct multiple GRU models, Meta learner ensemble predictions numerous trained models. We investigate our model on China Meteorological Administration (CMA) dataset compare results other non-ensemble-based techniques. Results show an apparent reduction model’s mean square error variance. code available GitHub: TC prediction.