Harshvardhan Pandey
Harshvardhan Pandey
@vprusso can I work on this?
@vprusso I have a small question regarding the already implemented strategies. Is it always guaranteed that the optimal measurement operators are hermitian? From what I understand, the measurement port just...
Thanks! That clears it up. One question regarding the SDP  This has simultaneous maximization of N functions p(rho_k|M_k) for k in range 1...N. It is possible that each of...
I see. So it is essentially maximize the maximum of p(rho_k|M_k)?
Interesting. That doesn't seem like a concave objective though. Because even if for each k, p(pho_k|M_k) is a concave function then their maximum is not guaranteed to be.
Yeah. Even I haven't been able to look into the paper properly yet. I'll do that.
From what it looks like based on section 2.5.2, for a given k, we can maximize p(rho_k|M_k) by just tuning M_k and then adjust M_{N+1} at end accordingly.
Hi @vprusso! Apologies for not getting back sooner, I was a little busy. I did try the solution mentioned in the paper. It works for the example they have given...
I was considering that we could just abandon the solution framework presented by the paper. In the end, we have N independent optimization problems for M_1, ..., M_N. Those can...
@purva-thakre it is just a simple rearrangement. For example in this case, we want to maximize trace(M_k * rho_k)/trace(M_k * rho). I can instead maximize trace(M_k * rho_k) with the...