Talk:Cross-entropy method

Untitled

For the arg max on line 3 of the Generic CE algorithm pseudocode, what is u? Should the u instead be v?

That is, should $\mathop {\textrm {argmax}} _{\mathbf {v} }{\frac {1}{N}}\sum _{i=1}^{N}H(\mathbf {X} _{i}){\frac {f(\mathbf {X} _{i};\mathbf {u} )}{f(\mathbf {X} _{i};\mathbf {v} ^{(t-1)})}}\log f(\mathbf {X} _{i};\mathbf {v} )$ be changed to $\mathop {\textrm {argmax}} _{\mathbf {v} }{\frac {1}{N}}\sum _{i=1}^{N}H(\mathbf {X} _{i}){\frac {f(\mathbf {X} _{i};\mathbf {v} )}{f(\mathbf {X} _{i};\mathbf {v} ^{(t-1)})}}\log f(\mathbf {X} _{i};\mathbf {v} )$

--Tekhnofiend (talk) 14:54, 2 April 2008 (UTC)

I can't believe that the pseudo-code is efficient; finding the optimal choice of the sampled points, i.e., the maximal one, seems much easier than sorting the array of all sampled points, yet the pseudo-code does the sort instead of finding the maximum.

--Allandon (talk) 15:09, 6 April 2011 (UTC)

@Allndon What is it your objection? You sounded like making a personal comment, not an objective constructive suggestion, This is a well-established technique with sound peer-review references. I am removing the banner. mcyp (talk) 01:01, 27 November 2017 (UTC)