Building a Deep Learning Model to Facilitate Software Project Risk Assessment
Ramakrishna Kolikipogu, Parag Rastogi, D. Victorseelan, Pavitar Parkash Singh · 2023
Software companies strive to create software projects of superior quality by leveraging the most optimal global resources at the most competitive cost. We implemented a methodology to attain global software development (GSD). Utilizing a method that involves collaborating on projects across various geographically dispersed locations, commonly called distributed development, is recommended. Companies encounter multiple challenges when they endeavour to implement Global Software Development (GSD) due to the inherent characteristics of GSD and its distinctions from conventional methodologies. Constructing a structure for machine learning to aid technology project risk analysis. The study's results indicate that 15 key factors significantly impact software projects in the context of GSD. Experimental analysis reveals that logistic regression and random forest models yield the most favourable outcomes, achieving 88% and 80% accuracy, respectively. Both models have a 71% and 72% area under the curve