Overview: PCA Models and Issues
Overview: PCA Models and Issues
复制标题
DOI：
10.1007/978-3-319-65558-1_1
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
R. Fernández;P. Louis;F. Nardi
中科院分区：
文献类型：
--
作者：
R. Fernández;P. Louis;F. Nardi
概率元胞自动机（PCA）是一种相互作用的离散随机动力系统，被用作广泛的自然和社会现象的建模工具。其主要特点是：（i）一个随机的组件，区分他们从众所周知的细胞自动机（CA）算法和（ii）一个潜在的并行性，使他们除了纯粹的异步模拟动力学统计力学，如相互作用的粒子系统和格劳伯动力学。在应用方面，这些功能使PCA成为高性能计算，分布式计算和仿真的一个有吸引力的计算框架。事实上，PCA已被很好地用于研究自然系统或大型互联网络结构的多尺度仿真框架的一部分。在数学方面，PCA具有丰富的数学理论，可以更好地理解随机性和同步性在大型系统演化中的作用。这本书是一个试图提出一个广泛的全景目前的地位PCA理论和应用。贡献涵盖重要的问题和应用概率，统计力学，计算机科学，自然科学和动力系统。这一章既是本书的指南，也是对本书所讨论问题的介绍。本章从PCA建模的一般概述开始，然后介绍了在不同背景下的显着应用。报告最后讨论了今后发展的方法和前景之间的联系。
Probabilistic cellular automata (PCA) are interacting discrete stochastic dynamical systems used as a modeling tool for a wide range of natural and societal phenomena. Their key features are: (i) a stochastic component that distinguishes them from the well-known cellular automata (CA) algorithms and (ii) an underlying parallelism that sets them apart from purely asynchronous simulation dynamics in statistical mechanics, such as interacting particle systems and Glauber dynamics. On the applied side, these features make PCA an attractive computational framework for high-performance computing, distributed computing, and simulation. Indeed, PCA have been put to good use as part of multiscale simulation frameworks for studying natural systems or large interconnected network structures. On the mathematical side, PCA have a rich mathematical theory that leads to a better understanding of the role of randomness and synchronicity in the evolution of large systems. This book is an attempt to present a wide panorama of the current status of PCA theory and applications. Contributions cover important issues and applications in probability, statistical mechanics, computer science, natural sciences, and dynamical systems. This initial chapter is intended both as a guide and an introduction to the issues discussed in the book. The chapter starts with a general overview of PCA modeling, followed by a presentation of conspicuous applications in different contexts. It closes with a discussion of the links between approaches and perspectives for future developments.