Renewable Energy Source Ranking and Analysis Using Fuzzy MCDM, ML, and XAI Techniques

To evaluate renewable energy alternatives under uncertainty, this study develops a hybrid decision support model combining fuzzy MCDM techniques with machine learning based prediction and XAI tools to enhance transparency in energy planning.Six renewable sources were assessed using Fuzzy TOPSIS and Fuzzy ELECTRE across six sustainability-related criteria: cost, efficiency, environmental impact, ease of implementation, technological innovation, and social acceptance.Entropy weighting was used to determine criterion importance.Fuzzy TOPSIS ranked hydropower highest (closeness coefficient = 0.7142) and wave energy lowest (0.3290).Fuzzy ELECTRE clarified dominance relationships among the options.For predictive analysis, machine learning models XGBoost, CatBoost, and Gradient Boosting were trained to forecast MCDM scores.XGBoost outperformed others (R = 0.9999, MAE = 0.0003).Explainable AI tools (SHAP and DALEX) revealed environmental impact and efficiency as the most influential factors.This integrated framework supports transparent and data-driven renewable energy planning and can inform sustainable policy decisions.