机器学习与量子场论
Machine learning with quantum field theories
摘要:
离散化的欧几里德场论和某类概率图模型(即马尔可夫随机场的数学框架)之间的精确等价性,为从量子场论的角度研究机器学习提供了机会。在这篇文章中,我们将证明,通过Hammersley-Clifford定理,$ϕ^{4}$标量场理论的正方形格满足当地的马尔可夫性质,因此可以作为一个马尔可夫随机场重铸。然后,我们将从$ϕ^{4}$理论推导出机器学习算法和神经网络,它们可以被视为传统神经网络架构的推广。最后,我们将总结基于$ϕ^{4}$机器学习算法的概率分布和目标概率分布之间的非对称距离最小化的应用。
英文摘要:
The precise equivalence between discretized Euclidean field theories and a certain class of probabilistic graphical models, namely the mathematical framework of Markov random fields, opens up the opportunity to investigate machine learning from the perspective of quantum field theory. In this contribution we will demonstrate, through the Hammersley-Clifford theorem, that the $ϕ^{4}$ scalar field theory on a square lattice satisfies the local Markov property and can therefore be recast as a Markov random field. We will then derive from the $ϕ^{4}$ theory machine learning algorithms and neural networks which can be viewed as generalizations of conventional neural network architectures. Finally, we will conclude by presenting applications based on the minimization of an asymmetric distance between the probability distribution of the $ϕ^{4}$ machine learning algorithms and target probability distributions.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。