Amortized Causal Discovery: Learning to Infer Causal Graphs from Time-Series Data

被引:0
作者
Lowe, Sindy [1 ]
Madras, David [2 ]
Zemel, Richard [2 ]
Welling, Max [1 ]
机构
[1] Univ Amsterdam, UvA Bosch Delta Lab, Amsterdam, Netherlands
[2] Univ Toronto, Vector Inst, Toronto, ON, Canada
来源
CONFERENCE ON CAUSAL LEARNING AND REASONING, VOL 177 | 2022年 / 177卷
基金
加拿大自然科学与工程研究理事会;
关键词
Causal Discovery; Granger Causality; Hidden Confounding; Noisy Observations; Amortization; Time-Series; Graph Neural Networks; MODELS;
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
On time-series data, most causal discovery methods fit a new model whenever they encounter samples from a new underlying causal graph. However, these samples often share relevant information which is lost when following this approach. Specifically, different samples may share the dynamics which describe the effects of their causal relations. We propose Amortized Causal Discovery, a novel framework that leverages such shared dynamics to learn to infer causal relations from time-series data. This enables us to train a single, amortized model that infers causal relations across samples with different underlying causal graphs, and thus leverages the shared dynamics information. We demonstrate experimentally that this approach, implemented as a variational model, leads to significant improvements in causal discovery performance, and show how it can be extended to perform well under added noise and hidden confounding.
引用
收藏
页数:17
相关论文
共 54 条
[1]  
Alet F., 2019, Advances in Neural Information Processing Systems (NeurIPS), V32, P11804
[2]   IDENTIFIABILITY OF PARAMETERS IN LATENT STRUCTURE MODELS WITH MANY OBSERVED VARIABLES [J].
Allman, Elizabeth S. ;
Matias, Catherine ;
Rhode, John A. .
ANNALS OF STATISTICS, 2009, 37 (6A) :3099-3132
[3]  
[Anonymous], 2012, Causality: Statistical perspectives and applications
[4]  
Fuchs FB, 2020, Arxiv, DOI arXiv:1907.12887
[5]  
Battaglia PW, 2016, ADV NEUR IN, V29
[6]  
Bengio Yoshua, 2019, INT C LEARN REPR
[7]  
Chickering D. M., 2003, Journal of Machine Learning Research, V3, P507, DOI 10.1162/153244303321897717
[8]  
Cremer C, 2018, PR MACH LEARN RES, V80
[9]  
Dhir A, 2020, AAAI CONF ARTIF INTE, V34, P3781
[10]  
Eichler Michael, 2012, Causal Inference in Time Series Analysis