IMPORTANT: the only thing truly worth looking at in this repo is Plots/point_1_posterior.png!
Building location prediction model from kaggle dataset the bayesian way.
Current state: Bayesian linear regression is used to convert signal to distance. Doughnuts of probabilty are built up around the senors. These can then be used to predict possible device positions.
To do: Current final combination is additive, it would be better/correct to use then product of probabilties.