This research study examines how ML techniques can assess computer science student performance. The study compares current and past performance metrics, identifies salient performance assessment features, and outlines strategies for sustaining or enhancing student achievement in the context of the COVID-19 pandemic, which has shifted the paradigm toward online learning. Context: With the rise of digital learning platforms and the changing educational environment, ML approaches are being used to get useful insights from massive student data sets. The COVID-19 pandemic has highlighted the need for improved virtual student learning monitoring and support methods. Objective: The purpose of this study is to determine if ML methods can accurately assess student computer science proficiency. Compare students' current performance measures to their prior records to identify COVID-19-related changes. To choose relevant characteristics for performance evaluation using ML algorithms and statistics. Methods: The research uses a mixed-methods approach, combining quantitative academic record analysis with qualitative student survey and interview data. Results: show substantial relationships between performance metrics and external influences, including remote learning. ML models identify academic performance factors, enabling targeted interventions and instructional improvements. Conclusion: Integrating ML technologies can improve computer science instruction and student achievement. Data-driven insights help instructors change teaching tactics, identify at-risk pupils, and create a dynamic learning environment that promotes academic success. ML techniques may help educational institutions negotiate distant learning and achieve inclusive, equitable, and successful results.