Tuesday, April 29, 2008

Clarification on HW6

Ok, yeah, part (b) was super-vague, sorry about that. Here is a more precise question, I think.

Consider the matrix M2(y); if this is psd, then we know, e.g., that yabcd can be written as the dot product of vectors zab and zcd, and also as the dot products of vectors zbc and zad etc. Similarly yab can be written as za dot zb, and also as zab dot z. Hence we now have these vectors z's, each subscripted by at most two vertices, with some constraints relating them.

The goal of the problem is to show that the SDP Q1(KVC) is at least as strong as a natural vertex cover SDP, which assigns a vector vi to each node i to minimize ivivi subject to
  • (v-va)(v-vb)=0 for all edges ab.
  • vv=1 (say).
Can you show that the SDP you've created implies these constraints?

Does your SDP also imply the triangle inequalities:
  • va-vb2+vb-vc2va-vc2?
If things are still vague (or if you see some mistakes above), please email/post yr questions.

No comments: