Automated defect identification via path analysis-based features with transfer learning

Zhang, Yuwei; Jin, Dahai; Xing, Ying; Gong, Yi

doi:10.1016/j.jss.2020.110585

Cited by 17 publications

(8 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The selection of these algorithms refers to previous studies, and they demonstrate the effectiveness of the algorithm in software defect prediction (Hall et al, 2012; Li, Jing, & Zhu, 2018; Wahono, 2015). When considering the implementation of the above‐mentioned algorithms, we hired Weka to complete this work (Witten et al, 2011), in which the default parameters provided by Weka were used in experiments, as mentioned in many studies (Menzies et al, 2015; Tantithamthavorn et al, 2016; Zhang, Jin, et al, 2020; Y. Zhang, Xing, et al, 2020). The mean of their experimental results is chosen as a benchmark for evaluating the performance of WBN and control approaches in CVDP and IVDP, ignoring their specificity.…”

Section: Methodsmentioning

confidence: 99%

Eliminating the high false‐positive rate in defect prediction through BayesNet with adjustable weight

et al. 2022

Self Cite

View full text Add to dashboard Cite

In defect prediction, a high false‐positive rate (FPR) caused by class imbalance not only increases the workload of testing and development but also consumes unnecessary costs. Many defect models against class imbalance have been proposed to improve the accuracy of defect prediction, but their ability to reduce FPR is unclear. To solve these problems, we first proposed a BayesNet with adjustable weights, called WBN, to reduce the FPR in software defect prediction, which is an algorithm independent of data preprocessing techniques. The mechanism of our WBN is to change the sampling probability of the misclassified instances when training the defect model, making the BayesNet model focus more on false alarm instances. And then, we investigate the FPR of five mainstream defect models for solving class imbalance and select them as comparison models to test the validity of our methods. The experimental result on eight open‐source projects shows that a) our WBN, in in‐version defect prediction (IVDP) and cross‐version defect prediction (CVDP), effectively reduces FPR with means of 0.384 and 0.322, respectively; b) compared with improved subclass discriminant analysis (ISDA) that is the lowest FPR in all control models, our WBN not only reduced the FPR but maintained recall whose mean value was 0.797, whereas ISDA did not, with an average recall of only 0.397; c) our WBN, in CVDP, not only reduces FPR, but also has significant superiority over five control defect models and baseline. Besides, we also found that the class imbalance difference between the test set and the training set has an impact on CVDP performance, recommending that practitioners choose the best dataset for CVDP from the defect data of the historical version through special technology.

show abstract

Section: Methodsmentioning

confidence: 99%

Eliminating the high false‐positive rate in defect prediction through BayesNet with adjustable weight

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…Summary of answers to RQ2: 178, 188, 189, 191, 193, 194, 196, 198, 201, 203, 204, 207ś213, 215, 218, 222, 223, 226, 227, 230, 232, 237, 239, 242, 248, 249, 253, 257, 259, 264, 271, 274, 276, 278ś280, 280, 282, 284, 285, 289, 290, 292, 294ś297, 299, 300, 303, 305, 306, 306, 307, 312, 319, 326, 330, 331, 336, 337] Area Under the Curve (AUC) The area under the receiver operating characteristics curve. Independent of the cutof value [15, 19, 29, 34, 43, 44, 46, 53, 56, [19,29,31,56,57,123,154,210,211,237,303,329,330] False positive rate(FPR) the ratio between the number of negative events wrongly categorized as positive and the total number of actual negative events + [32,152,163,163,164,198,221,…”

Section: Evaluation Metrics For Predictive Models In Classification Tasksmentioning

confidence: 99%

Predictive Models in Software Engineering: Challenges and Opportunities

Yang

Xia

et al. 2022

ACM Trans. Softw. Eng. Methodol.

View full text Add to dashboard Cite

Predictive models are one of the most important techniques that are widely applied in many areas of software engineering. There have been a large number of primary studies that apply predictive models and that present well-performed studies in various research domains, including software requirements, software design and development, testing and debugging and software maintenance. This paper is a first attempt to systematically organize knowledge in this area by surveying a body of 421 papers on predictive models published between 2009 and 2020. We describe the key models and approaches used, classify the different models, summarize the range of key application areas, and analyze research results. Based on our findings, we also propose a set of current challenges that still need to be addressed in future work and provide a proposed research road map for these opportunities.

show abstract

“…In recent years, automated sewer defects detection based on CNNs has been increasingly concerned, and achieved promising detection results [4,33,34]. The application of transfer learning technique further provides a state-of-the-art method to enhance model performance in terms of both detection accuracy and computational cost for structural defects detection [22,23]. A number of well-known pretrained CNNs have achieved success in the field of intelligent recognition thanks to their strong feature extraction abilities [35][36][37][38][39].…”

Section: Related Workmentioning

confidence: 99%

“…The application of transfer learning technique further provides a state-of-the-art method to enhance model performance in terms of both detection accuracy and computational cost for structural defects detection [22,23]. A number of well-known pretrained CNNs have achieved success in the field of intelligent recognition thanks to their strong feature extraction abilities [35][36][37][38][39]. Through retraining and finetuning the tailored CNNs as feature classifiers, the number of training samples and related computation time can be greatly saved when applied to new/novel tasks.…”

Section: Related Workmentioning

confidence: 99%

“…Furthermore, with aid of transfer learning (TL) [15,22,23], the model training time can be greatly shortened and the image quality can further be improved. However, to the best of our knowledge, previous studies on data augmentation in the field of sewer defects detection were mainly limited to the traditional methods, such as geometric and color transformations [4,22]. There is a lack of study utilizing StyleGANs to generate high-quality synthetic defects images for sewer systems.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Automated Sewer Defects Detection Using Style-Based Generative Adversarial Networks and Fine-Tuned Well-Known CNN Classifier

et al. 2021

View full text Add to dashboard Cite

Automated sewer defects detection has become an important trend for better management and maintenance of urban sewer systems. Deep learning technology has developed rapidly and offers an innovative solution for automated detection in engineering applications. However, insufficient data and unbalanced samples have proposed a big challenge to deep learning model training. This study adopts the state-of-the-art Style-based Generative Adversarial Networks (StyleGANs) model and compares the performances of its two variants in producing high-quality synthetic sewer defects images. Seven wellknown CNN models are further fine-tuned and trained using the synthetic images for automated sewer defects detection to examine the effects of StyleGANs on augmenting the detection performance. Results show that both StyleGANs are efficient in producing high-quality images with various styles and high-level details for multiple types of sewer defects. Specifically, the StyleGAN2-Adaptive Discriminator Augmentation (StyleGAN2-ADA) with the aid of Freeze Discriminator (Freeze-D) yields the best model performance. Among the adopted CNN classifiers, Inception_v3 achieves the highest detection accuracy. The mean detection accuracy is 94% (with a specific accuracy of 99.7%, 97%, 95.3% and 84% for tree root, residential wall, disjoint and obstacle, respectively) and confirms the reliability of the StyleGANs' performance. The study shows that StyleGANs provide a promising method to alleviate the limited and uneven dataset problem and can improve the deep learning model performance.

show abstract

Automated defect identification via path analysis-based features with transfer learning

Cited by 17 publications

References 18 publications

Eliminating the high false‐positive rate in defect prediction through BayesNet with adjustable weight

Eliminating the high false‐positive rate in defect prediction through BayesNet with adjustable weight

Predictive Models in Software Engineering: Challenges and Opportunities

Automated Sewer Defects Detection Using Style-Based Generative Adversarial Networks and Fine-Tuned Well-Known CNN Classifier

Contact Info

Product

Resources

About