I am reading a paper "A simpler approach to matrix completion", which talks about matrix completion using nuclear norm minimization. In section 4 it proofs that, the optimization problem gives a unique solution which is equal to the desired matrix.
It says that it uses dual feasibility approach to prove it, but I am not able to understand the proof at all. Can someone help?!
