Learning the Rules of Decision-Making: An Inverse Optimization Approach