Loading...
Please wait, while we are loading the content...
Similar Documents
Identifying Source Code Metrics to Improve Quality Predictive Models Using Genetic Algorithms
| Content Provider | Semantic Scholar |
|---|---|
| Author | Vivanco, Rodrigo |
| Copyright Year | 2006 |
| Abstract | In order to develop high quality software in a timely and costeffective manner, it is essential to identify and correct faulty components. Predictive models are used to classify modules as potentially faulty and in need of further inspection. Source code metrics can be used as input features for the classifier, however, there exists a large number of measures that capture different aspects of coupling, cohesion, inheritance, complexity and size. In large dimensional feature spaces some of the metrics may be irrelevant or redundant. Feature selection is the process of identifying a subset of features that improve the classifier's discriminatory performance. The focus of this study is to explore the efficacy of evolutionary algorithms to search for an optimum combination of object-oriented source code metrics to improve a predictive model's ability to identify fault-prone classes. |
| File Format | PDF HTM / HTML |
| Alternate Webpage(s) | http://www.cs.uoregon.edu/fse-14/docsym_docs/15_vivanco.pdf |
| Language | English |
| Access Restriction | Open |
| Content Type | Text |
| Resource Type | Article |
