Hi,
I am working on a binary recommendation system I want to predict whether an user buys a product or not, 0 fornot buying a product and 1 for buying it. I have gone through the (surprise) documentation and have set up the reader class with a rating_scale of 0 to 1 Reader(rating_scale=(0,1)), I have trained it using SVD and for the corresponding user id and product id I get a value between 0 and 1, so is it advisable to set some threshold like 0.5 or something(values less than it are consider as 0 and above or considered as 1)?Or is their any other method to approach this problem.
Thanks in advance.
Hi, I am working on a binary recommendation system I want to predict whether an user buys a product or not, 0 fornot buying a product and 1 for buying it. I have gone through the (surprise) documentation and have set up the reader class with a rating_scale of 0 to 1 Reader(rating_scale=(0,1)), I have trained it using SVD and for the corresponding user id and product id I get a value between 0 and 1, so is it advisable to set some threshold like 0.5 or something(values less than it are consider as 0 and above or considered as 1)?Or is their any other method to approach this problem. Thanks in advance.