Performance Analysis of Support Vector Machine Implementations on the D-Wave Quantum Annealer

In this paper a classical classification model, Kernel-Support Vector machine, is implemented as a Quadratic Unconstrained Binary Optimisation problem. Here, data points are classified by a separating hyperplane while maximizing the function margin. The problem is solved for a public Banknote Authentication dataset and the well-known Iris Dataset using a classical approach, simulated annealing, direct embedding on the Quantum Processing Unit and a hybrid solver. The hybrid solver and Simulated Annealing algorithm outperform the classical implementation on various occasions but show high sensitivity to a small variation in training data.