Ax=B Constrained Least Square Solver(s) and Performance

I am working on an Ax=B related problem, where both A and B are matrices, and x is a vector. My task is to solve for x given both A and B, and make sure the difference in Ax-B is as minimal as possible. Currently I am using the "quadprog" solver from the Optimization Toolbox because a) it supports sparse matrix, and b) I can constrain it to be within a fixed range, e.g., 0~255. While the solver is doing its job, I am wondering if there’s room to further improve the performance by perhaps adjusting the solver’s parameters. If this is the best ‘quadprog’ can do, are there any other solvers that can do even better?

Any input will be greatly appreciated. Thank you.

Answer 1

[Paul]

Anyone has any thoughts?