comparison toolboxes/distance_learning/mlr/loss/lossHingeDiagMKL.m @ 0:e9a9cd732c1e tip

first hg version after svn
author wolffd
date Tue, 10 Feb 2015 15:05:51 +0000
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:e9a9cd732c1e
1 function Xi = lossHingeDiagMKL(W, Psi, M, gradient)
2 %
3 % Xi = lossHingeDiagMKL(W, Psi, M, gradient)
4 %
5 % W: d*m matrix of diagonal metrics
6 % Psi: d*m feature matrix
7 % M: the desired margin
8 % gradient: if 0, returns the loss value
9 % if 1, returns the gradient of the loss WRT W
10
11 Xi = max(0, M - trace(W' * Psi));
12
13 if gradient & Xi > 0
14 Xi = -Psi;
15 end
16 end