Basic Statistical Estimation Outperforms Machine Learning in Monthly Prediction of Seasonal Climatic Parameters

Machine learning (ML) has been utilized to predict climatic parameters, and many successes have reported in the literature. In this paper, we scrutinize effectiveness of five widely used ML algorithms monthly prediction seasonal parameters using image data. Specifically, quantify predictive performance these applied various combinations features. We compare accuracy resulting trained models that basic statistical estimators are computed directly from training Our results show never significantly outperforms baseline, underperforms for most feature sets. Unlike previous similar studies, provide error bars relative different predictors based on jackknife estimates differences magnitudes. also practice shuffling data sequences which was employed some references leads leakage, over-estimated performance. Ultimately, paper demonstrates importance well-grounded techniques when producing analyzing models.