# Countywide gas consumption forecast
For my thesis, I've analyzed the problem of countywide gas consumption. I had data from a Croatian county that contained daily and hourly gas consumptions combined with weather data (temperature) and timestamps of the measurements. I've taken this data and applied various machine learning algorithms on it, trying to generate models that can correctly forecast gas consumptions several days ahead.
This research ended up used in a conference paper which can be read here (opens new window).