Codes for Hybrid CEEMDAN-Deep Learning Framework for Multi-Horizon Streamflow Forecasting under Climate Variability in the Kashmir Himalayas

This repository contains the MATLAB codes developed for the study: Hybrid CEEMDAN-Deep Learning Framework for Multi-Horizon Streamflow Forecasting under Climate Variability in the Kashmir Himalayas The scripts were used for:(1) Whale optimization algorithm (WOA) for feature selection and model ranking (2) Multivariate Copula analysis (3) Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN)(4) Multi-step ahead (MSA) river flow forecasting through hybrid CEEMDAN-MCMC-BC-BiLSTM-BiGRU model under direct, recursive, and direct-recursive forecasting strategy(5) Taylor diagram (6) SHAP Beeswarm Plot All codes were implemented in MATLAB.