Advanced Monte Carlo Methods for Inference in Complex Dynamic Models

用于复杂动态模型推理的高级蒙特卡罗方法

• 批准号: EP/K000276/1
• 负责人: Arnaud Doucet
• 金额: $ 78.97万
• 依托单位: University of Oxford
• 依托单位国家: 英国
• 项目类别: Fellowship
• 财政年份: 2013
• 资助国家: 英国
• 起止时间: 2013 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=EP%2FK000276%2F1
• 关键词: Advanced; Monte; Carlo; Methods; Inference

Many physical phenomena and much data can be accurately modelled using statistical or proba-bilistic models. Examples include the volatility of the stock market, gene expressions, radar signals,relational data or images. However, even if it is possible to obtain realistic physical models or satisfactory statistical models, only the simplest can be solved without the use of numerical methods.Examples of the need for such numerical methods include non-linear non-Gaussian time series models,Markov random fields, social networks models and so on. Thanks to the advent of enormous,cheap computational power and the development of a plethora of complex inference mechanisms, itis now possible, and in many real world systems it is becoming increasingly common, to employsophisticated simulation-based techniques to provide solutions to problems previously deemed insoluble. The intention behind the research program discussed herein is to extend current, and devisenovel, simulation-based architectures to attack and efficiently solve problems that are still deemedintractable.

许多物理现象和大量数据可以使用统计或概率模型进行精确建模。例子包括股票市场的波动性、基因表达、雷达信号、关系数据或图像。然而，即使可以获得现实的物理模型或令人满意的统计模型，如果不使用数值方法也只能解决最简单的问题。需要此类数值方法的例子包括非线性非高斯时间序列模型、马尔可夫随机场、社交网络模型等。由于巨大而廉价的计算能力的出现以及大量复杂推理机制的发展，现在可以采用复杂的基于模拟的技术来为以前认为无法解决的问题提供解决方案，而且在许多现实世界的系统中，这种情况变得越来越普遍。本文讨论的研究计划背后的目的是扩展当前的、设计新颖的基于模拟的架构，以攻击和有效地解决仍然被认为棘手的问题。

项目成果

Fast Computation of Wasserstein Barycenters

• 发表时间: 2013-10
• 作者: Marco Cuturi;A. Doucet

The Bouncy Particle Sampler: A Nonreversible Rejection-Free Markov Chain Monte Carlo Method

• DOI: 10.1080/01621459.2017.1294075
• 发表时间: 2018-01-01
• 期刊: JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION
• 影响因子: 3.7
• 作者: Bouchard-Cote, Alexandre;Vollmer, Sebastian J.;Doucet, Arnaud
• 通讯作者: Doucet, Arnaud

On the utility of Metropolis-Hastings with asymmetric acceptance ratio

论接受率不对称的 Metropolis-Hastings 的效用

• DOI: 10.48550/arxiv.1803.09527
• 发表时间: 2018
• 作者: Andrieu C

Generalized Polya Urn for Time-Varying Pitman-Yor Processes

用于时变 Pitman-Yor 过程的广义 Polya Urn

• 发表时间: 2017
• 期刊: JOURNAL OF MACHINE LEARNING RESEARCH
• 影响因子: 6
• 作者: Caron Francois

Distributed nonlinear consensus in the space of probability measures

• DOI: 10.3182/20140824-6-za-1003.00341
• 发表时间: 2014-04
• 期刊: IFAC Proceedings Volumes
• 作者: A. Bishop;A. Doucet