Prediction for Software Quality Open-Source Dataset Based on CART Approach Using Machine Learning Techniques

  • Krishan Kumar Department of Computer Science, Faculty of Technology, Gurukula Kangri (Deemed to be University), Haridwar, UK
Keywords: Classification tree, Regression tree, Fault prone module, and Quality prediction, J48, Random Forest and Logistic Model Tree.

Abstract

Open-source dataset available for different source of platform.Software quality is an important approach for the user as well as softwaredevelopers also.In this paper the study of various classification and regression tree (CART) method for software quality predictions. In this method have been used for a new algorithm design which is based on partitioning method. Among so many prediction methods over the recently published data mining predictionsfor software quality models such that classification and regression tree (CART), deep neural network (DNN), hierarchical attention network (HAN), multiple linear regression (MLR), stepwise regression (SR), artificial neural network (ANN) and case-based reasoning (CBR).CART algorithm to design an enhanced level of classification accuracy for large (complex) data set when compared with prior to classical CART concepts. CART can perform a balanced tree structure that is misclassify the fault-prone modules to classification-tree models based on Open-source dataset. The Open-source dataset used in Vote to classify the various machine learning algorithms such as J48, Random Forest, Logistic Model Tree. In this paper we also compute the best accuracy for vote dataset. This paper presents details on the cart algorithm to predict the quality of system in different ML algorithms.
Published
2021-02-25