A comparative study on the effect of data imbalance on software defect prediction

Liu, Yanbin; Zhang, Wen; Qin, Guangjie; Zhao, Jiangpeng

doi:10.1016/j.procs.2022.11.349

Cited by 11 publications

(2 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They exhibit variations in code size and include various software metrics. However, it is important to note that the NASA corpus is known to contain noisy attributes [17], have high dimensionality [18], and have imbalanced class records [19]. For example, the NASA JM1 dataset comprises 7,782 records with 1,672 containing defects and 6,110 without defects, each consisting of 22 attributes.…”

Section: Nasa Metrics Data Programmentioning

confidence: 99%

Optimized multi correlation-based feature selection in software defect prediction

Rahman,

Nugroho,

Faisal

et al. 2024

TELKOMNIKA

View full text Add to dashboard Cite

In software defect prediction, noisy attributes and high-dimensional data remain to be a critical challenge. This paper introduces a novel approach known as multi correlation-based feature selection (MCFS), which seeks to address these challenges. MCFS integrates two feature selection techniques, namely correlation-based feature selection (CFS) and correlation matrixbased feature selection (CMFS), intending to reduce data dimensionality and eliminate noisy attributes. To accomplish this, CFS and CMFS are applied independently to filter the datasets, and a weighted average of their outcomes is computed to determine the optimal feature selection. This approach not only reduces data dimensionality but also mitigates the impact of noisy attributes. To further enhance predictive performance, this paper leverages the particle swarm optimization (PSO) algorithm as a feature selection mechanism, specifically targeting improvements in the area under the curve (AUC). The evaluation of the proposed method is conducted on 12 benchmark datasets sourced from the NASA metrics data program (MDP) corpus, renowned for their noisy attributes, high dimensionality, and imbalanced class records. The research findings demonstrate that MCFS outperforms CFS and CMFS, yielding an average AUC value of 0.891, thereby emphasizing it is efficacy in advancing classification performance in the context of software defect prediction using k-nearest neighbors (KNN) classification.

show abstract

Section: Nasa Metrics Data Programmentioning

confidence: 99%

Optimized multi correlation-based feature selection in software defect prediction

Rahman,

Nugroho,

Faisal

et al. 2024

TELKOMNIKA

View full text Add to dashboard Cite

show abstract

“…In another work, Yanbin et al [6] investigated the impact of combining different sampling techniques and ML classifiers on defect prediction performance. While it finds no single optimal combination, it identifies support vector machines and deep learning as the most consistently performing classifiers.…”

mentioning

confidence: 99%

LCNN: Lightweight CNN Architecture for Software Defect Feature Identification Using Explainable AI

Begum,

Hasan Shuvo,

Kamal Nasir

et al. 2024

IEEE Access

View full text Add to dashboard Cite

A comparative study of software defect binomial classification prediction models based on machine learning

Tao,

Niu,

et al. 2024

Software Qual J

View full text Add to dashboard Cite

A comparative study on the effect of data imbalance on software defect prediction

Cited by 11 publications

References 27 publications

Optimized multi correlation-based feature selection in software defect prediction

Optimized multi correlation-based feature selection in software defect prediction

LCNN: Lightweight CNN Architecture for Software Defect Feature Identification Using Explainable AI

A comparative study of software defect binomial classification prediction models based on machine learning

Contact Info

Product

Resources

About