I want to solve a semi-definite problem with CVX
I have some pairs of similar points in a set called S
and some pairs of dissimilar points in a set called D
And I want to find a symmetric positive semi definite matrix by solving problem in below:
min_ A ∑_((x_i,x_j∈S)) ‖x_i-x_j ‖_A^2
s.t ∑_((x_i,x_j∈D)) ‖x_i-x_j ‖_A >= 1
A>=0
where ‖x_i-x_j ‖_A = √((x-y)^T A (x-y))
but I don’t know how to write it in a format that can be solved by CVX