The phase of the Software Development Life Cycle most vulnerable to and most important for mistakes is the process of forecasting potential risks contained within the program (SDLC). It could be crucial to the success of the project as a whole. Early risk prediction is essential for successful software development. In this paper, we propose a model for predicting software requirement dangers, using a pre-existing database of requirement risk information and state-of-the-art machine learning algorithms. Furthermore, many classifiers are compared to one another to determine the approach that will be best suitable for the model depending on the properties of the dataset. These results show that CDT is more effective than competing methods at reducing error rates. It is more likely that DT, DTNB, and CDT will succeed in their efforts to improve accuracy.

Accuracy, Dataset, Forecasting.Project, Software