Feature Selection Using Artificial Immune Network: An Approach for Software Defect Prediction

Software Defect Prediction (SDP) is a dynamic research field in the software industry. A quality product results customer satisfaction. However, higher number of user requirements, more complex will be software, with correspondingly probability failure. SDP challenging task requiring smart algorithms that can estimate component before it handed over to end-user. In this paper, we propose hybrid approach address particular issue. Our combines feature selection capability Optimized Artificial Immune Networks (Opt-aiNet) algorithm benchmark machine-learning classifiers for better detection bugs modules. proposed methodology was tested and validated using 5 open-source National Aeronautics Space Administration (NASA) data sets from PROMISE repository: CM1, KC2, JM1, KC1 PC1. Results were reported terms accuracy level an AUC highest accuracy, namely, 94.82%. The our experiments indicate improved by incorporating Opt-aiNet as (FS) method.