Overview / Usage

We use artificial intelligence to predict air pollution emergencies by associating weather variables to air quality, which basically a correlation based model. We also model the causal relationships between weather, wind flow, terrain and pollution emissions using a HPC mathematical model, which is a grid based differential equations. The challenge is to integrate both models.

Methodology / Approach

We have two alternatives. One is to integrate the results of the artificial intelligence model into the grid based differential equations model. The second one is to use artificial intelligence to improve the grid based model by selecting alternatives in design, execution flow and parameters. We intend to follow both paths.

Technologies Used

Data science: multifactor analysis, regression, classification

Deep learning: deep neural networks

OneAPI shared memory and distributed memory multiprocessing

GIS mapping

Documents and Presentations