Enhanced quantum state preparation via stochastic prediction of neural network

TL;DR

This paper introduces a stochastic prediction-based neural network method to enhance quantum state preparation in semiconductor quantum dots, improving efficiency and avoiding local optima during pulse sequence design.

Contribution

It presents a novel supervised learning approach leveraging stochastic predictions to optimize quantum state preparation, differing from reinforcement learning methods.

Findings

Improved quantum state preparation accuracy.

Reduced neural network complexity.

Enhanced optimization efficiency.

Abstract

In pursuit of enhancing the predication capabilities of the neural network, it has been a longstanding objective to create dataset encompassing a diverse array of samples. The purpose is to broaden the horizons of neural network and continually strive for improved prediction accuracy during training process, which serves as the ultimate evaluation metric. In this paper, we explore an intriguing avenue for enhancing algorithm effectiveness through exploiting the knowledge blindness of neural network. Our approach centers around a machine learning algorithm utilized for preparing arbitrary quantum states in a semiconductor double quantum dot system, a system characterized by highly constrained control degrees of freedom. By leveraging stochastic prediction generated by the neural network, we are able to guide the optimization process to escape local optima. Notably, unlike previous methodologies that employ reinforcement learning to identify pulse patterns, we adopt a training approach akin to supervised learning, ultimately using it to dynamically design the pulse sequence. This approach not only streamlines the learning process but also constrains the size of neural network, thereby improving the efficiency of algorithm.