Paper Key : IRJ************256
Author: Naimil Navnit Gadani
Date Published: 06 Jul 2024
Abstract
AI-based techniques have been developed and employed to detect improper use of the software or identify unexpected behaviors. In this context, AI-based methods have been used to find improper use of the system by analyzing the logs generated by the system. Similarly, AI has been used to find improper use by performing a statistical profile of use-cases or anomalies in the internal state of the system, i.e., detect runtime errors. Machine learning and deep learning are now commonly being applied to software development, monitoring, and maintenance to improve the software quality; i.e., AI can be used to construct a model to monitor how software is used in production and aid in the prediction of which code would introduce vulnerabilities. Model-based Bayesian deep learning technique is used for regression testing, where input entails the probability distribution range while output conveys how many probability distributions are compliant under such a range. 5AI is an umbrella term, encompassing both supervisedunsupervised learning approaches based on regressionclassification problems to an agent-based approach in which one would simulate human behavior. AI can analyze data to identify patterns and create models to make decisions (or predictions) based on probabilistic reasoning. AI can be used when establishing a combination of input attributes along with the needed output (discrete or continuous values) is difficult to obtain using the classical, algorithmic approach.
DOI Requested