No silver bullet in software analytics: Understanding the impact of model tuning metrics on the performance of software defects prediction models

Keyphrases

• Software Analytics 100%
• Software Defect Prediction 100%
• Model Tuning 100%
• Software Defect Prediction Model 100%
• Analytic Understanding 100%
• Model Evaluation 60%
• Inconsistency 40%
• Hyperparameter Optimization 40%
• Hyperparameter Tuning 40%
• Prediction Studies 40%
• Performance Improvement 20%
• Machine Learning Models 20%
• Training Data 20%
• Random Forest 20%
• Software Engineering 20%
• Software Project 20%
• Hyperparameters 20%
• Performance Metrics 20%
• K-nearest 20%
• Optimal Performance 20%
• Higher Rank 20%
• Data Complexity 20%
• Analytic Model 20%
• Engineering Problems 20%
• Historical Data 20%
• Class Overlap 20%
• Optimal Hyperparameters 20%
• Tree number 20%
• Ranking Model 20%
• MCC Score 20%
• Configurable Parameters 20%
• Multiple Tuning 20%
• Hyperparameter Selection 20%

Computer Science

• Prediction Model 100%
• Defect Prediction 100%
• Software Defect 100%
• Software Analytics 100%
• Leaning Parameter 42%
• Performance Model 28%
• Hyperparameter Optimization 28%
• Random Decision Forest 14%
• Software Engineering 14%
• Case Study 14%
• Software Project 14%
• Training Data 14%
• Good Indicator 14%
• Performance Metric 14%
• Performance Improvement 14%
• Optimal Performance 14%
• Reproducibility 14%
• Data Complexity 14%
• Historical Data 14%
• Analytic Model 14%
• Engineering Problem 14%
• Data Attribute 14%
• Machine Learning Model 14%
• Literature Review 14%