Year: 2020
Type: article
Abstract: Recently, machine learning algorithms using quantum computers have been actively developed. Among them, the support vector machine (SVM) specialized for the classification problem is one of the algorithms that are attracting the most attention for improving the performance by the quantum computer. In this study, we propose an SVM that classifies test data according to energy level by using time evolution calculation of Ising model. Furthermore, we propose a method to determine the support vector by applying the concept of DBSCAN and Deutsch-Jozsa algorithm and examining the mixture of surrounding data classes. We show that our algorithm can be classified regardless of the difference in the dimensions of the teacher data, and also the difference between linear and nonlinear data.
Institution: University of Electro-Communications
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W3000486968