Batch simulations and uncertainty quantification in Gaussian process surrogate approximate Bayesian computation

被引:0
作者
Jarvenpaa, Marko [1 ,2 ]
Vehtari, Aki [1 ]
Marttinen, Pekka [1 ]
机构
[1] Aalto Univ, Helsinki Inst Informat Technol HIIT, Dept Comp Sci, Espoo, Finland
[2] Univ Oslo, Dept Biostat, Oslo, Norway
来源
CONFERENCE ON UNCERTAINTY IN ARTIFICIAL INTELLIGENCE (UAI 2020) | 2020年 / 124卷
基金
芬兰科学院;
关键词
DESIGN; OPTIMIZATION; INFERENCE;
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The computational efficiency of approximate Bayesian computation (ABC) has been improved by using surrogate models such as Gaussian processes (GP). In one such promising framework the discrepancy between the simulated and observed data is modelled with a GP which is further used to form a model-based estimator for the intractable posterior. In this article we improve this approach in several ways. We develop batch-sequential Bayesian experimental design strategies to parallellise the expensive simulations. In earlier work only sequential strategies have been used. Current surrogate-based ABC methods also do not fully account the uncertainty due to the limited budget of simulations as they output only a point estimate of the ABC posterior. We propose a numerical method to fully quantify the uncertainty in, for example, ABC posterior moments. We also provide some new analysis on the GP modelling assumptions in the resulting improved framework called Bayesian ABC and discuss its connection to Bayesian quadrature (BQ) and Bayesian optimisation (BO). Experiments with toy and real-world simulation models demonstrate advantages of the proposed techniques.
引用
收藏
页码:779 / 788
页数:10
相关论文
共 51 条
  • [1] Acerbi L, 2018, ADV NEUR IN, V31
  • [2] Learning with Submodular Functions: A Convex Optimization Perspective
    Bach, Francis
    [J]. FOUNDATIONS AND TRENDS IN MACHINE LEARNING, 2013, 6 (2-3): : 145 - 373
  • [3] Beaumont MA, 2002, GENETICS, V162, P2025
  • [4] Approximate Bayesian computation with the Wasserstein distance
    Bernton, Espen
    Jacob, Pierre E.
    Gerber, Mathieu
    Robert, Christian P.
    [J]. JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY, 2019, 81 (02) : 235 - 269
  • [5] Probabilistic Integration: A Role in Statistical Computation?
    Briol, Francois-Xavier
    Oates, Chris J.
    Girolami, Mark
    Osborne, Michael A.
    Sejdinovic, Dino
    [J]. STATISTICAL SCIENCE, 2019, 34 (01) : 1 - 22
  • [6] Chai H. R., 2019, The 22nd International Conference on Artificial Intelligence and Statistics, P2751
  • [7] Bayesian experimental design: A review
    Chaloner, K
    Verdinelli, I
    [J]. STATISTICAL SCIENCE, 1995, 10 (03) : 273 - 304
  • [8] Ginsbourger D, 2010, ADAPT LEARN OPTIM, V2, P131
  • [9] Greenberg David, 2019, P 36 INT C MACH LEAR
  • [10] Gunter T, 2014, ADV NEUR IN, V27