Quantum Annealing for Optimization Problems

Abstract

Quantum computers are a new way that allow us to solve complex optimization problems that are intractable for classical computers. This bachelor thesis explores the use of quantum computers, more specifically, adiabatic quantum computing to optimize complex problems focusing on the Quantum Unconstrained Binary Optimization (QUBO) model. Through this model, quantum annealers can be an effective way of solving optimization problems. In this thesis we will study the viability of using quantum annealers to optimize neural networks as well as its precision and efficiency. In the work, we will develop a mathematical model to represent neural networks of any size, and with different activation functions in such a way that it can be used in quantum computer by using the QUBO model. We will also compare this method to other classical methods and see the benefits and downsides of using quantum annealers for this specific optimization task.