Optimisation of the adaptive neuro-fuzzy inference system for adjusting low-cost sensors PM concentrations

Martina Casari, Piotr A. Kowalski, Laura Po

Abstract: Highlights•Low-cost PM2.5 data gathered from six SPS30 Sensirions co-located with a reference station.•Innovative methodology combining fuzzy logic with neural networks for air quality adjustment.•Comparative analysis between machine learning and ANFIS.•PM data exploration and model optimization, helping to improve the accuracy and reliability of ANFIS.