Added proxy improvement. I was running into issues with the squared error implementation, so went through the calculations that sklearn do to find the proxy, and figured that they are missing a few parenthesis in their calculations, so might be an idea to have a walkthrough somewhere of what we are doing.
Added proxy improvement. I was running into issues with the squared error implementation, so went through the calculations that sklearn do to find the proxy, and figured that they are missing a few parenthesis in their calculations, so might be an idea to have a walkthrough somewhere of what we are doing.