RESEP researcher Cobus Burger builds winning model for Uber Movement SANRAL Cape Town Challenge
RESEP’s Cobus Burger recently won the Uber Movement SANRAL Cape Town Challenge, beating 738 data scientists from around the world. The objective was to develop a model that could accurately predict road accidents at an exact time and place on Cape Town’s busy roads. Using historic road incident data as well as traffic data from the Uber Movement platform, the aim of the competition was to assist authorities in anticipating where best to allocate resources at a given time, therefore making the roads around Cape Town a safer place.
Speaking to Zindi, Burger described his model as “a single lightGBM model…” of which “almost all the explanatory power … came from the [traffic light] camera data. Historic data was used to approximate how many cars will be using each road segment for each hour of each day of the week. These forecasts were highly correlated with traffic incidents. I also added daily deviations from these underlying patterns, although these did not add much. Weather and holiday data added less lift than I expected but were kept in the model, nonetheless.”
In second and third places were teams from Russia and Nigeria respectively.
Read more about the competition as well as interviews with the winner and runners up here.