Abstract:This paper provides an in-depth study and analysis of software defect prediction methods in a cloud environment and uses a deep learning approach to justify software prediction. A cost penalty term is added to the supervised part of the deep ladder network; that is, the misclassification cost of different classes is added to the model. A cost-sensitive deep ladder network-based software defect prediction model is proposed, which effectively mitigates the negative impact of the class imbalance problem on defect prediction. To address the problem of lack or insufficiency of historical data from the same project, a flow learning-based geodesic cross-project software defect prediction method is proposed. Drawing on data information from other projects, a migration learning approach was used to embed the source and target datasets into a Gaussian manifold. The kernel encapsulates the incremental changes between the differences and commonalities between the two domains. To this point, the subspace is the space of two distributional approximations formed by the source and target data transformations, with traditional in-project software defect classifiers used to predict labels. It is found that real-time defect prediction is more practical because it has a smaller amount of code to review; only individual changes need to be reviewed rather than entire files or packages while making it easier for developers to assign fixes to defects. More importantly, this paper combines deep belief network techniques with real-time defect prediction at a fine-grained level and TCA techniques to deal with data imbalance and proposes an improved deep belief network approach for real-time defect prediction, while trying to change the machine learning classifier underlying DBN for different experimental studies, and the results not only validate the effectiveness of using TCA techniques to solve the data imbalance problem but also show that the defect prediction model learned by the improved method in this paper has better prediction performance.

Bridging Effort-Aware Prediction and Strong Classification

Unifying Defect Prediction, Categorization, and Repair by Multi-Task Deep Learning

Improving Classifier-Based Effort-Aware Software Defect Prediction by Reducing Ranking Errors

Combined Classifier for Cross-Project Defect Prediction: an Extended Empirical Study.

Defect Category Prediction Based on Multi-Source Domain Adaptation

Within-Project Defect Prediction

A Compositional Model for Cross-Project Effort-Aware Defect Prediction

Effort-Aware semi-Supervised just-in-Time defect prediction

Investigating Associative Classification for Software Fault Prediction: an Experimental Perspective.

Deep Learning for Just-In-Time Defect Prediction

Deep Learning Software Defect Prediction Methods for Cloud Environments Research

Software Defect Prediction and Localization with Attention-Based Models and Ensemble Learning

Software defect prediction based on nested-stacking and heterogeneous feature selection

On effort-aware metrics for defect prediction

Deep Semantic Feature Learning with Embedded Static Metrics for Software Defect Prediction

Effort-aware and just-in-time defect prediction with neural network

Software defect prediction: future directions and challenges

Decision-Level Defect Prediction Based on Double Focuses

Software visualization and deep transfer learning for effective software defect prediction

Using Coding-Based Ensemble Learning to Improve Software Defect Prediction.

Predicting defects in object-oriented software using cost-sensitive classification