The frontier of simulation-based inference

被引:498
作者
Cranmer, Kyle [1 ,2 ]
Brehmer, Johann [1 ,2 ]
Louppe, Gilles [3 ]
机构
[1] NYU, Ctr Cosmol & Particle Phys, New York, NY 10003 USA
[2] NYU, Ctr Data Sci, New York, NY 10011 USA
[3] Univ Liege, Montefiore Inst, B-4000 Liege, Belgium
基金
美国国家科学基金会;
关键词
statistical inference; implicit models; likelihood-free inference; approximate Bayesian computation; neural density estimation; APPROXIMATE BAYESIAN COMPUTATION; SEQUENTIAL MONTE-CARLO;
D O I
10.1073/pnas.1912789117
中图分类号
O [数理科学和化学]; P [天文学、地球科学]; Q [生物科学]; N [自然科学总论];
学科分类号
07 ; 0710 ; 09 ;
摘要
Many domains of science have developed complex simulations to describe phenomena of interest. While these simulations provide high-fidelity models, they are poorly suited for inference and lead to challenging inverse problems. We review the rapidly developing field of simulation-based inference and identify the forces giving additional momentum to the field. Finally, we describe how the frontier is expanding so that a broad audience can appreciate the profound influence these developments may have on science.
引用
收藏
页码:30055 / 30062
页数:8
相关论文
共 74 条
[1]  
Andreassen A., 2019, ARXIV190708209V1
[2]  
[Anonymous], 2019, MON NOT R ASTRON SOC, DOI DOI 10.1093/MNRAS/STZ1960
[3]  
[Anonymous], 2019, JHEP, DOI DOI 10.1007/JHEP01(2019)057
[4]  
[Anonymous], **DATA OBJECT**, DOI DOI 10.5281/ZEN0D0.1634428)(2018)
[5]  
[Anonymous], 2019, COMPUT PHYS COMMUN, DOI DOI 10.1016/J.CPC.2019.06.007
[6]  
[Anonymous], 2018, STAT COMPUT, DOI DOI 10.1007/S11222-017-9738-6
[7]  
[Anonymous], 2003, P NATL ACAD SCI USA, DOI DOI 10.1073/PNAS.0306899100
[8]  
[Anonymous], 1984, ANN STAT
[9]  
[Anonymous], 2014, UNC ART INT
[10]  
[Anonymous], 2012, STAT COMPUT, DOI DOI 10.1007/S11222-012-9315-Y