Predicting Software Cohesion Metrics with Machine Learning Techniques

The cohesion value is one of the important factors used to evaluate software maintainability. However, measuring a relatively difficult issue when tracing source code manually. Although there are many static analysis tools, not every tool measures metric. user should apply different tools for metrics. In this study, besides use these we predicted values (LCOM2, TCC, LCC, and LSCC) with machine learning techniques (KNN, REPTree, multi-layer perceptron, linear regression (LR), support vector machine, random forest (RF)) solve them alternatively. We created two datasets utilizing open-source projects. According obtained results, LCOM2 TCC metrics, KNN algorithm provided best LCC LSCC REPTree was best. out all RF, had close performances each other, therefore any can be metric prediction.