Comparing the performance of machine learning algorithms using estimated accuracy

Abstract

In this paper, we have worked on comparing various data mining algorithms using R tool and various comparison models. After comparison has been done, we have applied the best algorithm as per the result to make the prediction. In this paper, we worked on how to check algorithms on a dataset using R and find the most accurate algorithm model for our dataset. It cannot be easy to tell what algorithm to use on the dataset to get the best results. We don't know the best parameters to use for particular algorithms. Here we worked on the strategy of trial and error to choose the accurate algorithm. Data set is used to train models and a test option is used to evaluate the model. Test metrics are used for comparison. Worked on the various models for which model to choose, to configure them, and pre-process them using data. Applied various techniques for comparing the accuracy of constructed models. We have worked on some algorithms compared them and after choosing one algorithm we can improve the result of algorithms by tuning various algorithms parameters by combining or changing parameters. Once we find the best algorithm applied on the dataset for prediction and tried to improve it by changing various criteria. Here in this paper, we have worked on the decision tree and tried to find out the best-resulting model for prediction. This learning opportunities can be further used in affordable energy, agriculture, and environmentally sound technologies etc.

Keywords

• Data mining
• Decision tree
• Dataset
• R Language
• SVM