Finding Probably Approximate Optimal Solutions by Training to Estimate the Optimal Values of Subproblems

Open in new window