Comparative Analysis of Weather Dataset Using Decision Tree, Random Forest and Gradient Boosted Tree Classification Algorithms in Cloud Computing

Abstract

This study presents a comparative analysis of weather prediction accuracy using three machine learning classification algorithms: Decision Tree (DT), Random Forest (RF), and Gradient Boosted Tree (GBT). The experiments were conducted on a weather dataset within a cloud computing environment, specifically using an Amazon Web Services (AWS) Elastic Compute Cloud (EC2) instance, simulating a real-world deployment. The models were evaluated based on key performance metrics: accuracy, precision, recall, F1-score, and computation time. The results demonstrate that GBT achieved the highest performance across all metrics, followed 
by RF and DT. However, while GBT and RF provided superior accuracy, they 
exhibited higher computational costs compared to DT, which was more 
computationally efficient but showed lower accuracy. The scalability of the models was also tested by increasing the dataset size, revealing that the decision tree scaled more efficiently than the ensemble-based models. This analysis provides valuable insights into the trade-offs between computational efficiency and predictive accuracy in cloud-based weather forecasting applications.