Research Paper | Computer Science & Engineering | India | Volume 2 Issue 5, May 2013
A Hybrid Approach of Module Sequence Generation using Neural Network for Software Architecture
Manjot Kalsi, Janpreet Singh
Software Engineering Principals gave a new direction to software industry . With the advent of software engineering principals we are now able to check the feasibility of modules under consideration which further will lead to restrict the happening of software crisis . But the selection of development of modules from the bunch of modules identified is totally dependent on the past experience of the project manager of project planner. For the detection of fault prone modules which will lead to software crisis if we are not implementing then in a desired order, various predictive models can be used based on source code metrics as input for the classifiers . The identification of metric subset for the enhancement of performance for predictive objective would not only improve the model but also provides strength to the structural properties of the modules. Unbalanced datasets also itself is a kind of difficulty for building predictive modeling. Unbalanced datasets are common in empirical software engineering as a majority of the modules are not likely to be faulty [8, 28]. We propose a method of applying search based metric selection and oversampling of NASA dataset . This paper is the extension of our previous review paper. In this proposed method the selection approach uses the weights of Neural Network to identify the sequence of implementation of software modules.
Keywords: Metric subset, Predictive models, unbalanced datasets, NASA dataset, Neural Network
Edition: Volume 2 Issue 5, May 2013
Pages: 133 - 137
How to Cite this Article?
Manjot Kalsi, Janpreet Singh, "A Hybrid Approach of Module Sequence Generation using Neural Network for Software Architecture", International Journal of Science and Research (IJSR), https://www.ijsr.net/search_index_results_paperid.php?id=IJSRON2013942, Volume 2 Issue 5, May 2013, 133 - 137