Software Defect Detection

Using dataset from Promise Repository (http://promise.site.uottawa.ca/SERepository/datasets-page.html) to prepare an ML model capable of detecting whether a software is defective. This repository contains datasets, which contain data from McCabe and Halstead features extractors of source code.

Approach

• Implemented Random Forests
• Logistic Regression
• Deep Neural Network
• 1D CNN

Also used PCA as a method to reduce dimensionality and enhance the accuracy