Optimization in Ensemble Model for Weather Prediction

The daily lives and livelihood of human beings and society are greatly affected and controlled by changes in weather, and prediction of the same becomes a useful exercise. Climate changes are becoming more erratic in recent times making the traditional methods less effective, and hence more modern approaches of weather prediction are the need of the hour. In this paper, a machine learning approach has been employed to predict the weather of a dataset available in the Kaggle forum. An extreme gradient boosting (XGBoost) algorithm has been developed and utilized to classify the weather data amongst 50 different weather conditions. A Bayesian optimization technique is applied to enhance the accuracy of the XGBoost. The results obtained indicate that the accuracy of the proposed models is superior to other algorithms.