Inspired by the world's move towards a greener future we wanted to see if we could predict CO2 emissions in a way that was effective and simple. We then open-sourced the simple code on GitHub.
Using Google's Tensorflow machine learning library we built a linear regression model. The system takes the history of the world's annual CO2 measurements and concludes a prediction for the next number in the sequence. We were pleased to hear that one year after the experiment the Mauna Loa Observatory, Hawaii confirmed the prediction.