Software risk prediction by Grasshopper Optimization Algorithm and Machine learning

Software development can be considered an activity that uses various technological advances and requires high knowledge. For this reason, every software development project contains elements of uncertainty known as project risk. The success of a software development project is highly dependent on the amount of risk associated with each project activity. Therefore, as a project manager, awareness of risks is not enough. In order to achieve a successful outcome, a project manager must be able to identify, then evaluate, prioritize, and ultimately manage all major risks.Risk management is a process to identify, eliminate and predict possible risks. In other words, this process is related to all the activities that are done to reduce the uncertainty associated with specific tasks or events. Risk management focuses on identifying risks and dealing with them appropriately. Projects have individual or general risks. Some of these risks are related to a specific activity and others to the project. Usually, risks are first identified and associated with project activities. Determining how people behave to achieve strategic activity goals is to identify risks. One of the methods of improvement is to take help from new algorithms and use machine learning techniques in the process of identifying or predicting possible risks.The use of different algorithms and techniques to identify software risks has always attracted the attention of experts. In this study, the grasshopper optimization algorithm has been used to improve classification accuracy and increase accuracy as well as reduce specificity. In this method, by combining the proposed algorithm with the support vector machine algorithm, it is used to obtain better and more acceptable results in data with large dimensions in order to reduce the dimension and also select the feature, which has attracted the attention of many researchers.The grasshopper optimization algorithm has not been used in the risk prediction system of software projects, the obtained results have shown that this proves the applicability and strength of the grasshopper optimization algorithm for solving problems with unknown and real search spaces.The purpose of this study is to predict the risks of software projects with the help of the grasshopper optimization algorithm. In this method, feature selection and reduction is done by the grasshopper optimization algorithm, and vector machine classification methods are used to classify risk and features.