Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Russian Physics Journal

Equivalence of Gibbs ensembles for quantum systems obeying Maxwell-Boltzmann statistics

V. D. Kondrashov, G. I. Nazin

Published Jul 1, 1984• 0 citations
Publisher
DOI
10.1007/bf00897442

Related papers

Quantum particle systems subjected to Maxwell-Boltzmann statistics under extremal conditions

V. D. Kondrashov

Russian Physics Journal•Apr 1, 1983•0 citations

Superstable interactions in quantum statistical mechanics : Maxwell-Boltzmann statistics

R. Esposito, Francesco Nicoló, Mario Pulvirenti

Annales De L Institut Henri Poincare-physique Theorique•Jan 1, 1982•6 citations

Bose-Einstein versus Maxwell-Boltzmann distributions

P. Enders, Ahornallee Senzig

The reply by Sands & Dunning-Davies [3] to my paper [1] is acknowledged concerning the distribution functions. Keywords: Maxwell-Boltzmann distribution, Bose-Einstein distribution, classical ensembles, state functions In my paper ‘Are there physical systems obeying the Maxwell-Boltzmann statistics?’[1] I have written down the well-known probabilities of the 4 possible results of one fair toss of 2 fair coins according to the Maxwell-Boltzmann (MB), Bose-Einstein (BE) and Fermi-Dirac (FD) statistics. The two results coin 1 = head, coin 2 = tail and coin 2 = head, coin 1 = tail are counting as different in MB and as one and the same in BE and in FD. Then, I have quoted Gibbs [2], that the interchange of two © 2011 C. Roy Keys Inc. — http://redshift.vif.comApeiron, Vol. 18, No. 1, January 2011 16

•Jan 1, 2011•1 citation

Machine learning algorithms based on generalized Gibbs ensembles

Tatjana Puškarov, Axel Cortés Cubero

Abstract Machine learning algorithms often take inspiration from the established results and knowledge from statistical physics. A prototypical example is the Boltzmann machine algorithm for supervised learning, which utilizes knowledge of classical thermal partition functions and the Boltzmann distribution. Recently, a quantum version of the Boltzmann machine was introduced by Amin et al , however, non-commutativity of quantum operators renders the training process by minimizing a cost function inefficient. Recent advances in the study of non-equilibrium quantum integrable systems, which never thermalize, have lead to the exploration of a wider class of statistical ensembles. These systems may be described by the so-called generalized Gibbs ensemble (GGE), which incorporates a number of ‘effective temperatures’. We propose that these GGEs can be successfully applied as the basis of a Boltzmann-machine–like learning algorithm, which operates by learning the optimal values of effective temperatures. We show that the GGE algorithm is an optimal quantum Boltzmann machine: it is the only quantum machine that circumvents the quantum training-process problem. We apply a simplified version of the GGE algorithm, where quantum effects are suppressed, to the classification of handwritten digits in the MNIST database. While lower error rates can be found with other state-of-the-art algorithms, we find that our algorithm reaches relatively low error rates while learning a much smaller number of parameters than would be needed in a traditional Boltzmann machine, thereby reducing computational cost.

arXiv:1804.03546•Journal of Statistical Mechanics Theory and Experiment•Oct 1, 2018•7 citations

Statistical Ensembles in Quantum Mechanics

D. I. Blokhint︠s︡ev

Springer eBooks•Jan 1, 1976•3 citations

Boltzmann, Gibbs, and Darwin-Fowler Approaches in Parastatistics

Roberto Leon Ponczek, C. C. Yan

There are essentially three methods used in equilibrium statistical mechanics, namely, the Boltzmann method of identifying the equilibrium sta te with the most probable one; the Gibbs ensemble method of postulating a canonical distribution; and the Darwin-Fowler method' of identifying the equilibrium state with the average state. Most textbooks on statis tical mechanics include discussions of the first two methods. The in clusion into the text of the last method, which schrudinger2 seems to advocate, is less usual . The equivalence of those methods is usual 1 y demonstrated for the i deal Boltzmann, Bose - Einstein, and Fermi-Dirac sta tistics. However, the equivalence is not guaranteed when the methodsare applied to different statistics such as parastatistics3, where eachener gy leve1 can afford up to some number k of particles, and can be regar ded as a generalization of the above mentioned statistics. It seems desirable to confirm t he equivalence of those methods for parastatistics. In most textbooks however discussions on parastatistics are r ather rare.

Revista brasileira de fisica•Jan 1, 1976•0 citations
The reply by Sands & Dunning-Davies [3] to my paper [1] is acknowledged concerning the distribution functions. Keywords: Maxwell-Boltzmann distribution, Bose-Einstein distribution, classical ensembles, state functions In my paper ‘Are there physical systems obeying the Maxwell-Boltzmann statistics?’[1] I have written down the well-known probabilities of the 4 possible results of one fair toss of 2 fair coins according to the Maxwell-Boltzmann (MB), Bose-Einstein (BE) and Fermi-Dirac (FD) statistics. The two results coin 1 = head, coin 2 = tail and coin 2 = head, coin 1 = tail are counting as different in MB and as one and the same in BE and in FD. Then, I have quoted Gibbs [2], that the interchange of two © 2011 C. Roy Keys Inc. — http://redshift.vif.comApeiron, Vol. 18, No. 1, January 2011 16
Abstract Machine learning algorithms often take inspiration from the established results and knowledge from statistical physics. A prototypical example is the Boltzmann machine algorithm for supervised learning, which utilizes knowledge of classical thermal partition functions and the Boltzmann distribution. Recently, a quantum version of the Boltzmann machine was introduced by Amin et al , however, non-commutativity of quantum operators renders the training process by minimizing a cost function inefficient. Recent advances in the study of non-equilibrium quantum integrable systems, which never thermalize, have lead to the exploration of a wider class of statistical ensembles. These systems may be described by the so-called generalized Gibbs ensemble (GGE), which incorporates a number of ‘effective temperatures’. We propose that these GGEs can be successfully applied as the basis of a Boltzmann-machine–like learning algorithm, which operates by learning the optimal values of effective temperatures. We show that the GGE algorithm is an optimal quantum Boltzmann machine: it is the only quantum machine that circumvents the quantum training-process problem. We apply a simplified version of the GGE algorithm, where quantum effects are suppressed, to the classification of handwritten digits in the MNIST database. While lower error rates can be found with other state-of-the-art algorithms, we find that our algorithm reaches relatively low error rates while learning a much smaller number of parameters than would be needed in a traditional Boltzmann machine, thereby reducing computational cost.
There are essentially three methods used in equilibrium statistical mechanics, namely, the Boltzmann method of identifying the equilibrium sta te with the most probable one; the Gibbs ensemble method of postulating a canonical distribution; and the Darwin-Fowler method' of identifying the equilibrium state with the average state. Most textbooks on statis tical mechanics include discussions of the first two methods. The in clusion into the text of the last method, which schrudinger2 seems to advocate, is less usual . The equivalence of those methods is usual 1 y demonstrated for the i deal Boltzmann, Bose - Einstein, and Fermi-Dirac sta tistics. However, the equivalence is not guaranteed when the methodsare applied to different statistics such as parastatistics3, where eachener gy leve1 can afford up to some number k of particles, and can be regar ded as a generalization of the above mentioned statistics. It seems desirable to confirm t he equivalence of those methods for parastatistics. In most textbooks however discussions on parastatistics are r ather rare.