A steady state quantum classifier

TL;DR

This paper introduces a quantum classifier based on a single qubit interacting with environments, exploiting quantum dynamical maps to achieve non-linear data response and linear separation of quantum data, with potential implementation in superconducting circuits.

Contribution

It proposes a novel open quantum classifier utilizing steady state responses and quantum dynamical maps, demonstrating linear separability of quantum data.

Findings

Steady state response exhibits non-linear behavior with input parameters.

Quantum data instances can be linearly separated.

Model is feasible with superconducting circuits under experimental conditions.

Abstract

We report that under some specific conditions a single qubit model weakly interacting with information environments can be referred to as a quantum classifier. We exploit the additivity and the divisibility properties of the completely positive (CP) quantum dynamical maps in order to obtain an open quantum classifier. The steady state response of the system with respect to the input parameters was numerically investigated and it's found that the response of the open quantum dynamics at steady state acts non-linearly with respect to the input data parameters. We also demonstrate the linear separation of the quantum data instances that reflects the success of the functionality of the proposed model both for ideal and experimental conditions. Superconducting circuits were pointed out as the physical model to implement the theoretical model with possible imperfections.