Calendar of Events
Using machine learning to identify toxic gases emitted from vehicle exhausts
Speaker: Unab Javed (Rutgers University)
There are over a billion vehicles in the world today. The laws that regulate their tail-pipe emissions grow more stringent each year, as governments attempt to control air pollution and carbon dioxide emissions. In this talk I will describe our recent work in creating a machine learning pipeline that is able to classify gases present in a gas mixture and predict their concentrations with high accuracy. The algorithm takes sensor readings as input and uses support vector machines to classify the gases in the gas mixture and then employs Gaussian process regression to predict gas concentrations. I will give an introduction to these two methods and then discuss our results. Sensor arrays combined with our algorithm can accurately determine the gases present in the vehicle exhausts and quantify them.