Should we implement ridge regression here?

[andcastillo]

Ridge regression is the most general algorithm for regression. It generalises all the regression methods implemented in this library as the problem of finding W such that the error e = (|| y-Wx ||) is minimised. Although it computational cost could be a little bigger for the already existing methods, It cover all the dimensionalities and all the function families:

[https://en.wikipedia.org/wiki/Tikhonov_regularization]

[targos]

As discussed today, yes if we want to implement it, it should be here