jackapbutler/ugandan-air-quality-detection

Using an ExtraTreesRegressor and Dense Neural Network to predict air quality (as measured by PM2.5 µ/m3) in Kampala, Uganda. # Ugandan-Air-Quality - AirQo competition to predict air quality (as measured by PM2.5 µ/m3) in Kampala, Uganda. - Finished in top 50 places globally with an ExtraTreesRegressor model in scikit-learn. # Problem Overview Until recently there has been a lack of data on air quality across sub-Saharan Africa. Reference grade monitors are extremely expensive and without access to data it is very difficult to raise awareness of the issues, or for government, business and individuals to know which actions to take to improve air quality and protect community health. AirQo has built a low-cost network of sensors and collected data across 65 locations in Uganda. These sensors provide them with a variety of weather and geographical data. # Key Variable PM2.5 One of the main indicators of air quality is PM2.5 (particulate matter smaller than 2.5 micrometers in diameter or around 1/30th the thickness of a human hair). These particles can be generated by traffic exhaust, industry, burning of fossil fuels and many other sources. The particles are so small that they are invisible to the naked eye and when inhaled do not just collect in the lungs and cause respiratory disease but can also enter the bloodstream and contribute to heart disease and stroke.