Risk Assessment and Management using Machine Learning Approaches

Appropriate estimation of risk and its management is a significant part in the lifecycle of software engineering. Risks are the crucial factors that affect the project's growth and work. The unidentified risk may damage the crucial components of the software. If the risk is underestimated, it may compromise the development and testing of the software. Looking at the facts, risk management stands as the crucial phase of any software development lifecycle. Risk management and assessment include various activities like recognizing, analyzing, planning, and controlling events that threaten project development. Machine learning, a booming area of research, has also shown its dominance in the field of risk assessment, however, it is still in the preliminary phase and has limited applicability. This paper considers various machine learning classifiers, such as Naive Bayesian, Decision Tree, Artificial Neural Network, and K Nearest Neighbor, and shows their comparative analysis and usage in developing the software risk assessment estimator. In order to boost the usage of machine learning models in the field of risk assessment, based on the findings, future directions are provided for researchers and stakeholders.