Unsupervised Deep Domain Adaptation for Heterogeneous Defect Prediction

Abstract

Heterogeneous defect prediction (HDP) is to detect the largest number of defective software modules in one project by using historical data collected from other projects with different metrics. However, these data can not be directly used because of different metrics set among projects. Meanwhile, software data have more non-defective instances than defective instances which may cause a significant bias towards defective instances. To completely solve these two restrictions, we propose unsupervised deep domain adaptation approach to build a HDP model. Specifically, we firstly map the data of source and target projects into a unified metric representation (UMR). Then, we design a simple neural network (SNN) model to deal with the heterogeneous and class-imbalanced problems in software defect prediction (SDP). In particular, our model introduces the Maximum Mean Discrepancy (MMD) as the distance between the source and target data to reduce the distribution mismatch, and use the cross-entropy loss function as the classification loss. Extensive experiments on 18 public projects from four datasets indicate that the proposed approach can build an effective prediction model for heterogeneous defect prediction (HDP) and outperforms the related competing approaches.

On the use of deep learning in software defect prediction

“…The fact that a large variety of different projects, versions and features is used in SDP leads to highly heterogeneous data, in particular when using different source and target for prediction. Such data degrade the performance of the classifier (Albahli, 2019, Gong et al, 2019Qiu et al, 2019a;Qiu et al, 2019b;Sheng et al, 2020;Sun et al, 2020a;Huang et al, 2021;Sun et al, 2021;Wu et al, 2021). Some researchers have tackled this challenge using different DL architectures which take this difference into account, while others have introduced normalization and transformation steps in data preprocessing as well as in feature extraction.…”

“…Hence, they proposed to learn a high-level feature representation from a bug dataset consisting of 19 mobile applications for JIT defect prediction Zeng et al (2021). used DL to build models for identifying defective commits in both WPDP and CPDP settings.Three studies addressed HDP Gong et al (2019). designed a neural network to deal with heterogeneous metric sets for defect prediction Sun et al (2021).…”

An optimized approach for class imbalance problem in heterogeneous cross project defect prediction

“…A simple Neural Network was being proposed for CPDP in the year 2019 to tackle HDP in which cross entropy function was applied for classification of error. 21 In the year 2021, researchers presented the work on semi-supervised learning for tackling heterogeneous defect prediction. The open-source projects were being utilized for the analysis.…”

Conditional Domain Adversarial Adaptation for Heterogeneous Defect Prediction

“…The number of neurons of output layer in generator is the number of metrics in target project, while the number of neurons of output layer in discriminator and classifier are one. In addition, based on method suggested by reference [54], we tune the parameter in the CDAA networks. Table 5 displays the hyper-parameter values set in our CDAA networks.…”