Closed olivierlabayle closed 3 years ago
I added a NonNegaqtiveLeastSquare interface to the https://github.com/ahwillia/NonNegLeastSquares.jl/ package. Need to see if they want to have it in the base package or not, see https://github.com/ahwillia/NonNegLeastSquares.jl/issues/33
Don't know how I feel about this issue. Loss functions are defined by the metalearner used. In the end it just means implementing a specific learner like the NNLL.