An Efficient Software Fault Prediction Scheme to Assure Qualified Software Implementation using Improved Classification Methods
Rajkumar N1, Viji C2
1Rajkumar N, Society for Vascular Surgery College of Engineering, Coimbatore (TamilNadu), India.
2Viji.C, Society for Vascular Surgery College of Engineering, Coimbatore (TamilNadu), India.
Manuscript received on 05 June 2019 | Revised Manuscript received on 12 June 2019 | Manuscript Published on 19 June 2019 | PP: 1-6 | Volume-8 Issue-8S June 2019 | Retrieval Number: H10010688S19/19©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Software quality is a main concern of software developers to ensure the required software that can provide user required services. However quality of software would be degraded considerably due to presence of the software faults in the programming language. Detection and removal of software faults requires more concern to be taken which needs to be concentrated more for improved performance. In the existing research work, accurate software fault prediction is done by introducing two stage data pre-processing stage which would select the more important features from the training data set and will result with the optimal training dataset thus the training accuracy can be improved. However existing research method doesn’t concentrate the dependencies between software modules and it doesn’t focus on the classification performance. These challenges are highlighted in the newly introduced research methodologies to obtain the accurate software prediction outcome by introducing the novel proposed research methodology namely Optimal and eliable Prediction of Software Faults (ORPSF). In the proposed research methodology, Optimal feature selection is performed by considering the inter relationship between the different features using Genetic Algorithm. This technique would select the optimal features which can detect the software faults accurately than the existing research methods. And the SVM classification approach is introduced to perform classification which can learn the training instances more accurately. Thus software fault prediction can be done accurately. The overall implementation of the proposed research technique is performed in the java simulation environment from which it can be shown that the proposed research methodology yields optimal results compared to the available research techniques.
Keywords: Software Fault Prediction, Feature Selection, Training Software Instances, Optimal Analysis, Software Quality.
Scope of the Article: Software Analysis, Design and Modelling