Can FTRL be applied on linear least squares? or is it just for logistic regression models? -
i'm exploring follow-the-regularized-leader ftrl proximal gradient descent: paper, reference implementation.
everywhere ftrl mentioned, loss surface gradient decent logloss
, , model prediction logistic regression
.
can use same algorithm linear least squares model? have problem want model linear model
, define loss least squares
, , ftrl find optimal solution - see problem that?
thanks.
i don't believe have studied enough ftrl i'm trying to. i've been doing research on algorithm , believe python code since uses least squares loss. since code written regression return value of 'predict' method changed since using sigmoid function regression useless. https://www.kaggle.com/scirpus/grupo-bimbo-inventory-demand/ftlr-use-pypy/code
i hope familiar python if not believe answer yes since i've tested same code on kaggle bimbo competition , produced similar score on test data 1 on public leader board.
Comments
Post a Comment