baptistar / bocs Goto Github PK

Hi,

I was just looking at the implementation of the Linear Regression model in BOCS/stat_model/LinReg.m and am a bit confused as to what is exactly happening in the LR.bayesTrain function. Would you mind providing more information on what this function does? To be specific, in lines 107-109, is this a Gaussian prior distribution on the parameters ? Next, how did you derive the posterior distribution for ? that you implemented between lines 117-119? Is this based on some source? If yes, would you mind pointing me to this source so I could learn more about this? Finally, what exactly are aPr and bPr taken as input to the function, and how are they updated in lines 125-126? Once again, a pointer to a source would be much appreciated!

Some BO repositories (e.g. GPyOpt, SMAC3) let the user set possible values of categorical variables and ranges of continuous variables. Is there a way to do that in BOCSpy?

Kindly add a link to your arxiv paper https://arxiv.org/abs/1806.08838 in the readme. Please refer to https://github.com/quark0/darts as an example of how and where the arxiv paper is to be linked. Thanks.

A Python 3 implementation will prove very useful. MATLAB is accessible to few due to its cost.