• Version
• Download 74
• File Size 431.31 KB
• File Count 1
• Create Date 6 June 2024
• Last Updated 6 June 2024

MACHINE LEARNING BASED SOFTWARE DEFECT PREDICTION USING PYTHON

Dr. U.NILABAR NISHA¹ , ARUNKUMAR  P S¹ , BABU S S², KISHOR M S³

^{SUBIKSHAN A S4}

^{`                       1}Head of the Department, Computer Science And Engineering, Mahendra Institute Of   Engineering And Technology, Namakkal-637503

²Student, Computer Science And Engineering, Mahendra Institute Of Engineering And Technology, Namakkal-637503

³Student, Computer Science And Engineering, Mahendra Institute Of Engineering And  Technology, Namakkal-637503

⁴Student, Computer Science And Engineering, Mahendra Institute Of Engineering And  Technology, Namakkal-63750

Abstract— Software development and the maintenance life cycle are lengthy processes. However, the possibility of having defects in the software can be high. Software reliability and performance are essential measures of software success, which affects user satisfaction and software cost. Predicting software defects using machine learning (ML) algorithms is one approach in this direction. Implementing this approach in the earlier stages of the software development improves software performance quality and reduces software maintenance cost. Different models and techniques have been implemented in many studies to predict software defects. This investigation implements ML algorithms, such as artificial neural networks (ANNs), random forest (RF), random tree (RT), decision table (DT), linear regression (LR), gaussian processes (GP), SMOreg, and M5P. A new software defect prediction model for software future defect prediction is proposed. The defect prediction is based on historical data. The results showed that a combination of ML algorithms could be used effectively to predict software defects. The SMOreg classifier scored the best performance results, where the ANN classifier scores the worst results.

Keywords— software defect prediction, software defect, prediction model, machine learning (ML), artificial neural networks (ANN), SMOreg Classifier.