8000 (Feature) Recommend top N items given a user · Issue #9 · gbolmier/funk-svd · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content
(Feature) Recommend top N items given a user #9
Open
@BradKML

Description

@BradKML

Querying every item on the table would be inefficient. There could be a way to make this process simpler by multiplying a user's vector with the whole item matrix.
Guessing: np.dot(self.pu_[u_ix], self.qi_) + self.bi_+ self.bu_[u_ix]

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions

      0