I did a research on Forest Fire Dataset which is available on UCI Machine Learning Repository.The main aim of the problem was to predict the burned area of a forest in Portugal according to different parameters like Temperature, Humidity and also by considering standard codes like DMC(Duff Moisture Code) and FFMC(Fine Fuel Moisture Code).It was a regression task but classification algorithms were also done in order to gain insgihts on the dataset.In regression simple linear regression was performed which gave the Mean square error of 0.14739.In order to perform classification many classification algorithms like Decision Trees,Neural Network,K Nearest Neighbours was performed and Decision Trees gave the best accuracy of 92.6%.