tag:blogger.com,1999:blog-961962165941544126.post7864641990562914795..comments2013-01-07T22:19:07.700-08:00Comments on Unpopped Kernels: Pop QuizAnonymoushttp://www.blogger.com/profile/15624261037800893086noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-961962165941544126.post-79753450027695703622012-12-13T19:32:27.941-08:002012-12-13T19:32:27.941-08:00Well I could apply the polar transformation to the...Well I could apply the polar transformation to the SVM program and also make it non-convex in $r$ and $\theta$. But SVM is convex in $w$ and $b$.Anonymoushttps://www.blogger.com/profile/15624261037800893086noreply@blogger.comtag:blogger.com,1999:blog-961962165941544126.post-8108509894017558402012-12-13T18:54:07.864-08:002012-12-13T18:54:07.864-08:00Hmm. you're computing the max of the lower env...Hmm. you're computing the max of the lower envelope of a collection of trig functions. not sure why it's EVER convex except on the line, where it reduces to the min of the x_i. Suresh Venkatasubramanianhttps://www.blogger.com/profile/15898357513326041822noreply@blogger.com