Validating models of quantum computer performance

Jincheng Zhang

This paper presents a novel framework for quantum computing model validation based on dynamic topology. The core idea is to leverage the inherent properties of dynamic topology to rigorously verify the correctness and efficiency of quantum algorithms, enhancing the reliability and performance of quantum computation. We explore the application of this framework to quantum information and quantum cryptography, demonstrating its potential to improve the robustness of these critical areas. This work contributes to the advancement of quantum algorithms and provides a new approach to validating the complex and sensitive nature of quantum systems.

Megan Dahlhauser, Robin Blume-Kohout, Timothy J Proctor +1 more

Malte Lochau, Ina Schaefer

Megan Dahlhauser, Kevin C Young, Timothy J Proctor +1 more

Highlights• Generalized for many common quantum characterization, verification, and validation models • Identify if a model is consistent with data, and if not, where the inconsistencies are • Explore trends of inconsistencies with various features, such as circuit depth or gate types • Report quantifiable differences among experiment, model, and target SummaryWe are developing methods to determine and report characterization performance on circuit prediction tasks using statistical tests and simulation.We are demonstrating this process in experiment using gate set tomography to evaluate performance on circuit prediction tasks.We aim to create a validation process that is an effective tool in identifying characterization shortcomings and upgrading methodologies to create more accurate characterizations of quantum devices.

Li, Jing, Gao, Fei, Lin, Song +5 more

Cross-validation is one of the important tools in machine learning, which is generally used for performance evaluation. It uses different portions of the data to test and train a model on different iterations, which leads to a high computational cost. In this paper, we present a quantum version of k-fold cross-validation to choose a good parameter for the nearest neighbor classification algorithm with a threshold t, where the classification performance is estimated efficiently. With the help of amplitude amplification and estimation, the proposed quantum algorithm achieves a polynomial speedup on the number of samples over its classical counterpart.