Inverse Parametric Optimization For Learning Utility Functions From Optimal and Satisficing Decisions