Comparative Analysis of Decision Tree, Random Forest, and X-Gradient Boost Algorithms for Weather Prediction in Seattle

Abstract

Weather prediction plays an essential role in supporting various sectors such as agriculture, transportation, and urban planning, particularly in regions with highly dynamic weather conditions like Seattle. This study aims to analyze and compare the performance of three machine learning algorithms Decision Tree, Random Forest, and X-Gradient Boost in predicting weather conditions using historical meteorological data. The dataset was obtained from Kaggle and includes several key attributes, such as precipitation, humidity, air pressure, temperature, and wind speed. The research methodology consists of data collection, preprocessing, model training, and performance evaluation using accuracy, precision, recall, and F1 score metrics. Model implementation and experimentation were conducted using the Google Colab platform, with hyperparameter tuning applied through GridSearchCV to optimize model performance. The experimental results indicate that the X-Gradient Boost algorithm achieved the highest accuracy of 84%, followed by Random Forest with 83.96% and Decision Tree with 83% after tuning. Based on these results, X-Gradient Boost is identified as the most effective algorithm for weather prediction in this study. These findings are expected to contribute to the development of more accurate and reliable machine learning-based weather forecasting systems.