Construction of Non-Convex Polynomial Loss Functions for Training a Binary Classifier with Quantum Annealing

   Abstract