A Learning-to-Rank Approach to Software Defect Prediction

Yang, Xiaoxing; Tang, Ke; Yao, Xin

doi:10.1109/tr.2014.2370891

Cited by 128 publications

(70 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We provide a comprehensive evaluation and comparison of the LTR approach against more algorithms for constructing SDP models for the ranking task. In previous work LTR approach can be compared with many other methods [17]. Linear Learning-to-rank approach gives better result as compare to count models.…”

Section: Learning-to-rank Approachmentioning

confidence: 99%

Performance Enhancement in Learning-to-Rank Approach for Software Defect Prediction

Priya¹,

Kaur²,

Sharma³

2017

IJSEIA

View full text Add to dashboard Cite

show abstract

Section: Learning-to-rank Approachmentioning

confidence: 99%

Performance Enhancement in Learning-to-Rank Approach for Software Defect Prediction

Priya¹,

Kaur²,

Sharma³

2017

IJSEIA

View full text Add to dashboard Cite

show abstract

“…Recently, Canfora et al [8,9] proposed the application of multi-objective genetic algorithms to generate a set of classification models considering file size and recall as two objectives to optimize. Yang et al [38] used GAs to optimize the ranks of defect-prone software components predicted by simple Linear Regression model (LR) without taking into account the inspection cost.…”

Section: Training Regression Models With Gasmentioning

confidence: 99%

“…Specifically, previous papers use GAs for calibrating algorithms to predict the defect proneness as a binary outcome (i.e., defective or non-defective artifacts) while we use regression models that, by definition, predict continuous values (e.g., number of defects) as done in [11] when measuring the cost-effectiveness. The most important difference with respect to previous approaches is that they use traditional performance metrics for classification problem as fitness functions to optimize [13,19,23,38]. As explained in Section 2, traditional performance are not wellsuited to evaluate prediction since they give the same priority/importance to all defect prone software components, independently from their size (cost) and the number of bugs (effectiveness).…”

Section: Training Regression Models With Gasmentioning

confidence: 99%

A Search-based Training Algorithm for Cost-aware Defect Prediction

Panichella

Alexandru

Panichella

et al. 2016

Proceedings of the Genetic and Evolutionary Computation Conference 2016

View full text Add to dashboard Cite

Research has yielded approaches to predict future defects in software artifacts based on historical information, thus assisting companies in effectively allocating limited development resources and developers in reviewing each others' code changes. Developers are unlikely to devote the same effort to inspect each software artifact predicted to contain defects, since the effort varies with the artifacts' size (cost) and the number of defects it exhibits (effectiveness). We propose to use Genetic Algorithms (GAs) for training prediction models to maximize their cost-effectiveness. We evaluate the approach on two well-known models, Regression Tree and Generalized Linear Model, and predict defects between multiple releases of six open source projects. Our results show that regression models trained by GAs significantly outperform their traditional counterparts, improving the cost-effectiveness by up to 240%. Often the top 10% of predicted lines of code contain up to twice as many defects. ABSTRACTResearch has yielded approaches to predict future defects in software artifacts based on historical information, thus assisting companies in effectively allocating limited development resources and developers in reviewing each others' code changes. Developers are unlikely to devote the same effort to inspect each software artifact predicted to contain defects, since the effort varies with the artifacts' size (cost) and the number of defects it exhibits (effectiveness). We propose to use Genetic Algorithms (GAs) for training prediction models to maximize their cost-effectiveness. We evaluate the approach on two well-known models, Regression Tree and Generalized Linear Model, and predict defects between multiple releases of six open source projects. Our results show that regression models trained by GAs significantly outperform their traditional counterparts, improving the cost-effectiveness by up to 240%. Often the top 10% of predicted lines of code contain up to twice as many defects.

show abstract

“…Based on the investigation of historical metrics [1][2], defect prediction aims to detect the defect proneness of new software modules. Therefore, defect prediction is often used to help to reasonably allocate limited development and maintenance resources [3][4][5]. With the advent of big data era and the development of machine learning techniques [6], many machine learning algorithms are applied to solve the practical problems in life [7][8][9].…”

Section: Introductionmentioning

confidence: 99%

A Data Filtering Method Based on Agglomerative Clustering

Yu¹,

Peipei²,

Zhang³

et al. 2017

International Conferences on Software Engineering and Knowledge Engineering

View full text Add to dashboard Cite

Abstract-Cross-company defect prediction (CCDP) is a practical way that trains a prediction model by exploiting one or multiple projects of a source company and then applies the model to target company. Unfortunately, larger irrelevant crosscompany (CC) data usually makes it difficult to build a crosscompany defect prediction model with high performance. To address such issues, this paper proposes a data filtering method based on Agglomerative Clustering (DFAC) for cross-company defect prediction. First, DFAC combines within-company instances and cross-company instances and uses Agglomerative clustering algorithms to group these instances. Second, DFAC selects sub-clusters which consist at least one WC instance, and collects the CC instances in the selected sub-clusters into a new CC data. Compared with existing data filter methods, the experimental results on 15 public PROMISE datasets show that DFAC increases PD value, reduces PF value and achieves higher G-measure and AUC values.

show abstract

A Learning-to-Rank Approach to Software Defect Prediction

Cited by 128 publications

References 24 publications

Performance Enhancement in Learning-to-Rank Approach for Software Defect Prediction

Performance Enhancement in Learning-to-Rank Approach for Software Defect Prediction

A Search-based Training Algorithm for Cost-aware Defect Prediction

A Data Filtering Method Based on Agglomerative Clustering

Contact Info

Product

Resources

About