Simularia AI Solutions
Artificial Intelligence Value Chain
Accurate and high spatial-temporal resolution concentrations maps of airborne pollutants are required in multiple fields (forecast, epidemiology, etc.) It is therefore required to increase the spatial resolution of traditional chemical transport models beyond the typical 1 km scale.
We apply a combined technique based on chemical transport models and ML models (Random Forest, XGBoost, etc.) The latter are trained with measurements from air quality monitoring stations thus improving both the accuracy and the spatial resolution of the concentration maps.
Healthcare / Social Services
Available on the market
Comparison of Machine Learning Models
Descent of scale from simulations with FARM model
A multi-city air pollution population exposure study
Combined use of chemical-transport and random-Forest models with dynamic population data