Summary

More mathematical approach to recommendation problem. No additional data about items and users is needed - just incomplete matrix.

License

CC-BY-SA 3.0

Tags

collaborative-filtering low-rank matrices matrix-completion recommendation

Feature Processing

Parameters

Operating System

independent

Code

%pseudocode

function norm = fobnorm(X) norm = sum(sum(X.*X)) end;

function err = tooptimize(X) err = fubnorm(X-Y); end;

function X = matrixcompletion(y,M,sizeX) % Matrix completion approach to collaborative filtering % finds: argmax_X ||X - Y||_p

% Output % X - matrix to be estimated % Input % y - sampled entries % M - masking operator, applied to vectorized form of X % sizeX - size of matrix to be reconstructed Y = reshape(M(y,2),sizeX);

X = oprimize(tooptimize)

end

Software Packages

Description

(No information yet)

URLs

(No information yet)

Publications

revision 1

by kidzik on 2011-09-13 20:16

No one has posted any comments yet. Perhaps you would like to be the first?

Leave a comment

This item was downloaded 0 times and viewed 4692 times.