Near-term quantum-classical associative adversarial networks

Anschuetz, Eric R.; Zanoci, Cristian

doi:10.1103/physreva.100.052327

Cited by 14 publications

(12 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Visualization may be another approach to compare classes of parameterized models for a range of problem instances in terms of, for example, the quality, number, or connectivity of local minima. This extends to quantum-classical algorithms where the aim is to enhance classical algorithms with a quantum component [51][52][53][54][55], as well as classical algorithms for physics, such as variational machine learning models for quantum states [56][57][58]. In this case, one could use the visualization techniques to evaluate the quality of their approximation to the true quantum state.…”

Section: Discussionmentioning

confidence: 99%

ORQVIZ: Visualizing High-Dimensional Landscapes in Variational Quantum Algorithms

Rudolph¹,

Sim²,

Raza³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Variational Quantum Algorithms (VQAs) are promising candidates for finding practical applications of near to mid-term quantum computers. There has been an increasing effort to study the intricacies of VQAs, such as the presence or absence of barren plateaus and the design of good quantum circuit ansätze. Many of these studies can be linked to the loss landscape that is optimized as part of the algorithm, and there is high demand for quality software tools for flexibly studying these loss landscapes. In our work, we collect a variety of techniques that have been used to visualize the training of deep artificial neural networks and apply them to visualize the high-dimensional loss landscapes of VQAs. We review and apply the techniques to three types of VQAs: the Quantum Approximate Optimization Algorithm, the Quantum Circuit Born Machine, and the Variational Quantum Eigensolver. Additionally, we investigate the impact of noise due to finite sampling in the estimation of loss functions. For each case, we demonstrate how our visualization techniques can verify observations from past studies and provide new insights. This work is accompanied by the release of the open-source Python package orqviz, which provides code to compute and flexibly plot 1D and 2D scans, Principal Component Analysis scans, Hessians, and the Nudged Elastic Band algorithm. orqviz enables flexible visual analysis of high-dimensional VQA landscapes and can be found at: github.com/zapatacomputing/orqviz.

show abstract

Section: Discussionmentioning

confidence: 99%

ORQVIZ: Visualizing High-Dimensional Landscapes in Variational Quantum Algorithms

Rudolph¹,

Sim²,

Raza³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…. 10) Similar to Eq. ( 1), D, G, and the associative memory are trained by optimizing the objective functions defined as follows:…”

Section: Associative Adversarial Networkmentioning

confidence: 99%

“…Several studies 10,11) have used quantum sampling methods when training the BM acting as the AAN's memory. One such approach ?)…”

Section: Introductionmentioning

confidence: 99%

“…17) It has been shown that the quantum AAN slightly outperforms the classical AAN when using the Gibbs sampler to draw samples from the BM for an MNIST 18) dataset-a large database of handwritten digits. Another approach using classical quantum sampling 10) has also shown that a quantum AAN works better than a classical AAN on the MNIST and cifar10 datasets (the cifar10 dataset consists of colored images). These results indicate that quantum sampling has a slightly positive effect on the GAN.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Comparing the effects of Boltzmann machines as associative memory in Generative Adversarial Networks between classical and quantum sampling

Urushibata,

Ohzeki,

Tanaka

2022

Preprint

View full text Add to dashboard Cite

We investigate the quantum effect on machine learning (ML) models exemplified by the Generative Adversarial Network (GAN), which is a promising deep learning framework. In the general GAN framework the generator maps uniform noise to a fake image. In this study, we utilize the Associative Adversarial Network (AAN), which consists of a standard GAN and an associative memory. Further, we set a Boltzmann Machine (BM), which is an undirected graphical model that learns low-dimensional features extracted from a discriminator, as the memory. Owing to difficulty calculating the BM's log-likelihood gradient, it is necessary to approximate it by using the sample mean obtained from the BM, which has tentative parameters. To calculate the sample mean, a Markov Chain Monte Carlo (MCMC) is often used. In a previous study, this was performed using a quantum annealer device, and the performance of the "Quantum" AAN was compared to that of the standard GAN. However, its better performance than the standard GAN is not well understood. In this study, we introduce two methods to draw samples: classical sampling via MCMC and quantum sampling via quantum Monte Carlo (QMC) simulation, which is quantum simulation on the classical computer. Then, we compare these methods to investigate whether quantum sampling is advantageous. Specifically, calculating the discriminator loss, the generator loss, inception score and Fréchet inception distance, we discuss the possibility of AAN. We show that the AANs trained by both MCMC and QMC are more stable during training and produce more varied images than the standard GANs. However, the results indicate no difference in sampling by QMC simulation compared to that by MCMC.

show abstract

“…Finally, after completing the work, a preprint was posted that covers similar ideas on quantum-classical associative adversarial networks that was performed independently from ours. In this work, the authors investigate a quantumclassical associative model and sample the latent space with quantum Monte Carlo (Anschuetz and Zanoci 2019).…”

Section: Contributionsmentioning

confidence: 99%

Quantum-assisted associative adversarial network: applying quantum annealing in deep learning

Wilson

Vandal

Hogg

et al. 2021

Quantum Mach. Intell.

View full text Add to dashboard Cite

Generative models have the capacity to model and generate new examples from a dataset and have an increasingly diverse set of applications driven by commercial and academic interest. In this work, we present an algorithm for learning a latent variable generative model via generative adversarial learning where the canonical uniform noise input is replaced by samples from a graphical model. This graphical model is learned by a Boltzmann machine which learns low-dimensional feature representation of data extracted by the discriminator. A quantum processor can be used to sample from the model to train the Boltzmann machine. This novel hybrid quantum-classical algorithm joins a growing family of algorithms that use a quantum processor sampling subroutine in deep learning, and provides a scalable framework to test the advantages of quantum-assisted learning. For the latent space model, fully connected, symmetric bipartite and Chimera graph topologies are compared on a reduced stochastically binarized MNIST dataset, for both classical and quantum sampling methods. The quantum-assisted associative adversarial network successfully learns a generative model of the MNIST dataset for all topologies. Evaluated using the Fréchet inception distance and inception score, the quantum and classical versions of the algorithm are found to have equivalent performance for learning an implicit generative model of the MNIST dataset. Classical sampling is used to demonstrate the algorithm on the LSUN bedrooms dataset, indicating scalability to larger and color datasets. Though the quantum processor used here is a quantum annealer, the algorithm is general enough such that any quantum processor, such as gate model quantum computers, may be substituted as a sampler.

show abstract

Near-term quantum-classical associative adversarial networks

Cited by 14 publications

References 67 publications

ORQVIZ: Visualizing High-Dimensional Landscapes in Variational Quantum Algorithms

ORQVIZ: Visualizing High-Dimensional Landscapes in Variational Quantum Algorithms

Comparing the effects of Boltzmann machines as associative memory in Generative Adversarial Networks between classical and quantum sampling

Quantum-assisted associative adversarial network: applying quantum annealing in deep learning

Contact Info

Product

Resources

About