Predicting and Enhancing Indoor Environmental Quality in Educational Environments: Latest Trends in AI and Machine Learning Applications

Indoor environmental quality (IEQ) affects students' health and academic performance who spend substantial time in school environments. This systematic literature review article explores how Artificial intelligence (AI) and Machine learning (ML) can predict and enhance IEQ within educational environments. A systematic review of 20 articles published between 2021 and 2024 was conducted using the PRISMA method, with all articles retrieved from the Scopus database. This article addresses three key research questions: 1) What IEQ factors are measured in university buildings, and why are they studied? 2) Which IoT sensors are employed in university buildings, and how are their data collected and managed? and 3) Which machine learning models are used for occupant comfort prediction in IEQ studies? The findings suggest air quality and thermal comfort are the most studied IEQ factors. These factors are primarily monitored using IoT sensors that enable continuous, real-time monitoring. None of the reviewed studies directly addressed acoustic comfort, highlighting a gap for future research. These IoT sensors are often integrate edge and cloud computing to ease data collection and reduce disruptions. Most articles used supervised learning models, of which 50% used deep learning techniques exclusively, 20% relied on traditional models, and 30% adopted a hybrid strategy combining traditional and deep learning ML. Reinforcement learning methods have also shown promise for dynamic HVAC and lighting control, although their adoption remains less common.