Comparative Analysis of Machine Learning Algorithms on Temperature and Wind Speed Prediction

Abstract

This research delves into weather forecasting, demonstrating the effectiveness of machine learning algorithms compared to traditional methods. Weather forecasts are pivotal in industries like agriculture, transportation, and disaster management, requiring precise predictions for parameters such as temperature, precipitation, humidity, and wind speed. The complexity and non-linear nature of meteorological data can challenge conventional techniques, impacting prediction accuracy. This project employs Random Forest and Multiple Linear Regression algorithms on historical data, mainly temperature and wind speed, collected from meteorological offices and dataset from Oikolab website. By analyzing data with MATLAB, the study aims to showcase machine learning's potential in improving short-term weather predictions, benefiting various industries and everyday planning.