Water demand, meteorological variables, pumping capacity

Missing value treatment, outlier correction, normalization

Descriptive statistics, CV, correlation matrix

Consumption, capacity, variability

(RT, RF, FUX, KSTAR, LWL)

Model Training and Testing

(Train-test split, grid search, cross-validation)

Model Evaluation and Comparison Statistical Metrics

(R², MAE, RMSE, MBE, NSE)

Demand Forecast & Risk Category for Each Station