Application of support vector machine models for forecasting solar and wind energy resources: A review