Comparative Performance Analysis of Machine Learning Classifiers on Weather Data

Accurate weather forecasting is essential for predicting the conditions of atmosphere based on time and location. It is used to determine the future climatic conditions of expectation. The Seattle weather data from Washington United States of America (U.S.A) is taken for this analysis and implemented using data mining tool for calculating and classifying its performance. In this study, the machine learning techniques are applied for classifying the weather forecast using various machine learning algorithms for giving the best accuracy. Based on variables, comparative performance is done using SMO, Random Forest and J48 algorithms. Thus, the weather data is classified with maximum accuracy of 98% using Random Forest compared with J48 and SMO algorithm. The suggested strategy improves that the rain weather data forecast is higher in Seattle region based on other weather conditions.