Software Defect Prediction Using Joint Harmony Search Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software defect prediction models suffer from reduced performance due to non-optimal parameter settings, particularly in preprocessing and classification model construction, leading to inefficient resource allocation and defect identification.
Innovation Solution
A method and device that utilize a cost-sensitive decision tree based on the harmony search algorithm to simultaneously optimize parameters in preprocessing, feature selection, class imbalance learning, and decision tree model construction, enhancing performance through adaptive parameter adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual parameter setting based on developer experience is used, then model configuration is simple, but prediction model performance is significantly reduced
Solution Approach 1:
The system performs self-optimization by automatically selecting optimal parameters for preprocessing and classification models using the harmony search algorithm, eliminating the need for manual developer intervention while achieving superior prediction performance
Solution Approach 2:
The invention systematically varies and optimizes multiple parameters including kernel types for SVM, K values for k-NN, preprocessing techniques, and feature selection methods to identify the optimal parameter combination that maximizes prediction accuracy
2Reliability
If separate optimization of preprocessing and classification parameters is performed, then each parameter can be optimized individually, but overall model performance is not maximized
Solution Approach 1:
The invention merges the optimization of preprocessing parameters and classification model parameters into a unified harmony search framework that simultaneously optimizes both sets of parameters, ensuring that their interactions are properly considered to achieve maximum overall performance
Solution Approach 2:
The harmony search algorithm serves as a universal optimization mechanism that handles multiple types of parameters (preprocessing, feature selection, and classification) within a single framework, making the system efficient and comprehensive
3Measurement precision
If optimal parameters are not adjusted, then model construction is faster, but defect prediction accuracy is significantly reduced
Solution Approach 1:
The system implements feedback through cross-validation performance evaluation, where the harmony search algorithm iteratively adjusts parameters based on prediction accuracy metrics and converges to the optimal parameter set that maximizes defect prediction performance
Data Source
AI summary
A performance enhancement method for a software defect prediction model according to one embodiment includes a software defect prediction model providing step of providing the software defect prediction model that identifies a module in which a software defect occurs, and a parameter optimization step of simultaneously optimizing at least one parameter in each step of a software defect prediction process by using an optimization algorithm to enhance performance of the software defect prediction model and in which a preprocessing step and a classification model generation step are simultaneously performed for a search space of the optimization algorithm.