From Detection to Solution: A Review of Machine Learning in PM2.5 Sensing and Sustainable Green Mitigation Approaches (2021–2025)

Particulate matter 2.5 (PM2.5) pollution poses severe threats to public health, ecosystems, and urban sustainability. With increasing industrialization and urban sprawl, accurate pollutant monitoring and effective mitigation of PM2.5 have become global priorities. Recent advancements in machine learning (ML) have revolutionized PM2.5 sensing by enabling high-accuracy predictions, and scalable solutions through data-driven approaches. Meanwhile, sustainable green technologies—such as urban greening, phytoremediation, and smart air purification systems—offer eco-friendly, long-term strategies to reduce PM2.5 levels. This review, covering research publications from 2021 to 2025, systematically explores the integration of ML models with conventional sensor networks to enhance pollution forecasting, pollutant source attribution, and intelligent pollutant monitoring. The paper also highlights the convergence of ML and green technologies, including nature-based solutions and AI-driven environmental planning, to support comprehensive air quality management. In addition, the study critically examines integrated policy frameworks and lifecycle-based assessments that enable equitable, sector-specific mitigation strategies across industrial, transportation, energy, and urban planning domains. By bridging the gap between cutting-edge technology and sustainable practices, this study provides a comprehensive roadmap for researchers to combat PM2.5 pollution.