Write a Blog >>
ICSE 2020
Wed 24 June - Thu 16 July 2020
Thu 9 Jul 2020 08:17 - 08:25 at Baekje - I16-Testing and Debugging 2 Chair(s): Rui Abreu

Defect prediction is an important task for preserving software quality. Most prior work on defect prediction uses software features, such as the number of lines of code, to predict whether a file or commit will be defective in the future. There are several reasons to keep the number of features that are used in a defect prediction model small. For example, using a small number of features avoids the problem of multicollinearity and the so-called `curse of dimensionality’. Feature selection and reduction techniques can help to reduce the number of features in a model. Feature selection techniques reduce the number of features in a model by selecting the most important ones, while feature reduction techniques reduce the number of features by creating new, combined features from the original features. Several recent studies have investigated the impact of feature \emph{selection} techniques on defect prediction. However, there do not exist large-scale studies in which the impact of multiple feature \emph{reduction} techniques on defect prediction is investigated.

In this paper [1], we study the impact of eight feature reduction techniques on the performance and the variance in performance of five supervised learning and five unsupervised defect prediction models. In addition, we compare the impact of the studied feature reduction techniques with the impact of the two best-performing feature selection techniques (according to prior work).

The following findings are the highlights of our study: (1) The studied correlation and consistency-based feature selection techniques result in the best-performing supervised defect prediction models, while feature reduction techniques using neural network-based techniques (restricted Boltzmann machine and autoencoder) result in the best-performing unsupervised defect prediction models. In both cases, the defect prediction models that use the selected/generated features perform better than those that use the original features (in terms of AUC and performance variance). (2) Neural network-based feature reduction techniques generate features that have a small variance across both supervised and unsupervised defect prediction models. Hence, we recommend that practitioners who do not wish to choose a best-performing defect prediction model for their data use a neural network-based feature reduction technique.

Thu 9 Jul

Displayed time zone: (UTC) Coordinated Universal Time change

08:05 - 09:05
I16-Testing and Debugging 2Technical Papers / Journal First at Baekje
Chair(s): Rui Abreu Instituto Superior Técnico, U. Lisboa & INESC-ID
08:05
12m
Talk
Low-Overhead Deadlock PredictionTechnical
Technical Papers
Yan Cai Institute of Software, Chinese Academy of Sciences, Ruijie Meng University of Chinese Academy of Sciences, Jens Palsberg University of California, Los Angeles
08:17
8m
Talk
The Impact of Feature Reduction Techniques on Defect Prediction ModelsJ1
Journal First
Masanari Kondo Kyoto Institute of Technology, Cor-Paul Bezemer University of Alberta, Canada, Yasutaka Kamei Kyushu University, Ahmed E. Hassan Queen's University, Osamu Mizuno Kyoto Institute of Technology
08:25
8m
Talk
The Impact of Correlated Metrics on the Interpretation of Defect ModelsJ1
Journal First
Jirayus Jiarpakdee Monash University, Australia, Chakkrit Tantithamthavorn Monash University, Australia, Ahmed E. Hassan Queen's University
08:33
8m
Talk
The Impact of Mislabeled Changes by SZZ on Just-in-Time Defect PredictionJ1
Journal First
Yuanrui Fan Zhejiang University, Xin Xia Monash University, Daniel Alencar Da Costa University of Otago, David Lo Singapore Management University, Ahmed E. Hassan Queen's University, Shanping Li Zhejiang University
08:41
8m
Talk
Which Variables Should I Log?J1
Journal First
Zhongxin Liu Zhejiang University, Xin Xia Monash University, David Lo Singapore Management University, Zhenchang Xing Australia National University, Ahmed E. Hassan Queen's University, Shanping Li Zhejiang University
08:49
12m
Talk
Understanding the Automated Parameter Optimization on Transfer Learning for Cross-Project Defect Prediction: An Empirical StudyTechnicalArtifact Available
Technical Papers
Ke Li University of Exeter, Zilin Xiang University of Electronic Science and Technology of China, Tao Chen Loughborough University, Shuo Wang , Kay Chen Tan City University of Hong Kong
Pre-print