This project is connected to an ongoing research project funded by the Swedish Transport Administration, and its focus is on the use of regression analysis to predict the travel time considering several types of input including road characteristics, weather, time of day, etc. The main input will be GPS-based travel time, and the project is expected to involve all of the steps of the machine learning pipeline, including data preparation, outlier detection, model formulation, etc. A special focus is expected to be on analysing how well the results generalize over different, sometimes geographically, separated road segments.
A prerequisite is courses in mathematical statistics and machine learning or similar.

Contact persons are Benny Lövström blo@bth.se (only during November), Henrik Fredriksson (hfn@bth.se) or Mattias Dahl (mdh@bth.se) at the Department of Mathematics and Natural Sciences, BTH. Note that these proposals are also published at other locations, so please take contact with us at an early stage to show your interest and check the availability.