Research Article
Cost-Sensitive Radial Basis Function Neural Network Classifier for Software Defect Prediction
Table 2
Attribute information of the datasets.
| Number | Attribute type | Quality metrics | Attribute definition |
| 1 | McCabe’s measure | loc | Line count of code | 2 | | Cyclomatic complexity | 3 | | Essential complexity | 4 | | Design complexity |
| 5 | Basic Halstead measures | loCode | Line count | 6 | loComment | Count of lines of comments | 7 | loBlank | Count of blank lines | 8 | loCodeAndComment | Count of code and comment lines | 9 | uniqOp | Unique operators | 10 | uniqOpnd | Unique operands | 11 | total_Op | Total operators | 12 | total_Opnd | Total operands | 13 | branchCount | Branch count of the flow graphs |
| 14 | Derived Halstead measures | | Total operators + operands | 15 | | Volume | 16 | | Program length | 17 | | Difficulty | 18 | | Intelligence | 19 | | Effort | 20 | | Estimate of the effort | 21 | | Time estimator |
| 22 | Output defect measure | Defects | false, true: module has/does not have one or more reported defects |
|
|