Multi-Class Classification Using Quantum Kernel Methods

Abstract

Quantum machine learning has recently attracted attention in various research fields. One of the most promising areas are kernel methods in quantum computers as they leverage the quantum computers advantage over classical kernels. The embedding of data in Hilbert space of quantum computers is called Quantum Embedding Kernels (QEKs). Many previous researches have explored the idea of using quantum embedding kernels for binary classification problems, demonstrating the advantage of quantum computing. This research is concerned with using these methods for multi-class classification problem and benchmark the results against well-known datasets such as IRIS and MNIST.