Trust Your $\nabla$: Gradient-based Intervention Targeting for Causal Discovery
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2022
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914738662801408 |
|---|---|
| author | Olko, Mateusz Zając, Michał Nowak, Aleksandra Scherrer, Nino Annadani, Yashas Bauer, Stefan Kuciński, Łukasz Miłoś, Piotr |
| author_facet | Olko, Mateusz Zając, Michał Nowak, Aleksandra Scherrer, Nino Annadani, Yashas Bauer, Stefan Kuciński, Łukasz Miłoś, Piotr |
| contents | Inferring causal structure from data is a challenging task of fundamental importance in science. Observational data are often insufficient to identify a system's causal structure uniquely. While conducting interventions (i.e., experiments) can improve the identifiability, such samples are usually challenging and expensive to obtain. Hence, experimental design approaches for causal discovery aim to minimize the number of interventions by estimating the most informative intervention target. In this work, we propose a novel Gradient-based Intervention Targeting method, abbreviated GIT, that 'trusts' the gradient estimator of a gradient-based causal discovery framework to provide signals for the intervention acquisition function. We provide extensive experiments in simulated and real-world datasets and demonstrate that GIT performs on par with competitive baselines, surpassing them in the low-data regime. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2211_13715 |
| institution | arXiv |
| publishDate | 2022 |
| record_format | arxiv |
| spellingShingle | Trust Your $\nabla$: Gradient-based Intervention Targeting for Causal Discovery Olko, Mateusz Zając, Michał Nowak, Aleksandra Scherrer, Nino Annadani, Yashas Bauer, Stefan Kuciński, Łukasz Miłoś, Piotr Machine Learning Artificial Intelligence Methodology Inferring causal structure from data is a challenging task of fundamental importance in science. Observational data are often insufficient to identify a system's causal structure uniquely. While conducting interventions (i.e., experiments) can improve the identifiability, such samples are usually challenging and expensive to obtain. Hence, experimental design approaches for causal discovery aim to minimize the number of interventions by estimating the most informative intervention target. In this work, we propose a novel Gradient-based Intervention Targeting method, abbreviated GIT, that 'trusts' the gradient estimator of a gradient-based causal discovery framework to provide signals for the intervention acquisition function. We provide extensive experiments in simulated and real-world datasets and demonstrate that GIT performs on par with competitive baselines, surpassing them in the low-data regime. |
| title | Trust Your $\nabla$: Gradient-based Intervention Targeting for Causal Discovery |
| topic | Machine Learning Artificial Intelligence Methodology |
| url | https://arxiv.org/abs/2211.13715 |