# Active Invariant Causal Prediction: Experiment Selection through Stability

@article{Gamella2020ActiveIC, title={Active Invariant Causal Prediction: Experiment Selection through Stability}, author={Juan L Gamella and Christina Heinze-Deml}, journal={arXiv: Methodology}, year={2020} }

A fundamental difficulty of causal learning is that causal models can generally not be fully identified based on observational data only. Interventional data, that is, data originating from different experimental environments, improves identifiability. However, the improvement depends critically on the target and nature of the interventions carried out in each experiment. Since in real applications experiments tend to be costly, there is a need to perform the right interventions such that as…

## 23 Citations

### Learning Neural Causal Models with Active Interventions

- Computer ScienceArXiv
- 2021

This work introduces an active intervention targeting (AIT) method which enables a quick identification of the underlying causal structure of the data-generating process and is applicable for both discrete and continuous optimization formulations of learning the underlying directed acyclic graph (DAG) from data.

### Near-Optimal Multi-Perturbation Experimental Design for Causal Structure Learning

- Computer ScienceNeurIPS
- 2021

This paper addresses the largely unexplored problem of designing experiments that simultaneously intervene on multiple variables, and develops efficient algorithms for optimizing different objective functions quantifying the informativeness of experiments.

### Operationalizing Complex Causes: A Pragmatic View of Mediation

- Computer ScienceICML
- 2021

It is demonstrated, on a range of simulated and real-world-inspired examples, that the proposed approach allows us to efficiently estimate the effect of crude interventions with limited data from new treatment regimes.

### visible and invisible : causal variable learning and its application in a cancer study

- Computer Science
- 2020

This work proposes a novel ε-ICP algorithm, based upon the promising invariant causal prediction (ICP) framework, that is implemented with a typed neural-symbolic functional language and demonstrated the application on a real-world cancer study dataset, Swiss CRC.

### Towards Out-Of-Distribution Generalization: A Survey

- Computer ScienceArXiv
- 2021

This paper serves as the first effort to systematically and comprehensively discuss the OOD generalization problem, from the definition, methodology, evaluation to the implications and future directions.

### Towards IID representation learning and its application on biomedical data

- Computer ScienceArXiv
- 2022

By utilizing the representation obtained via the learned function that induces IID, prediction of molecular characteristics on two biomedical datasets with real-world distribution shifts introduced by a) preanalytical variation and b) sampling protocol is conducted.

### Learning to Induce Causal Structure

- Computer ScienceArXiv
- 2022

The performance of the model improves as it observes more interventions, this suggest that the model is able to extract useful information from interventions in order to predict the causal structure.

### Interventions, Where and How? Experimental Design for Causal Models at Scale

- Computer ScienceArXiv
- 2022

This work incorporates recent advances in Bayesian causal discovery into the Bayesian optimal experimental design framework, allowing for active causal discovery of large, nonlinear SCMs while selecting both the interventional target and the value.

### Differentiable Multi-Target Causal Bayesian Experimental Design

- Computer Science
- 2023

This work proposes a conceptually simple end-to-end gradient-based optimization procedure to acquire a set of optimal intervention target-state pairs and demonstrates that this method outperforms baselines and existing acquisition strategies in both single-target and multi-target settings across a number of synthetic datasets.

### LEA: Latent Eigenvalue Analysis in application to high-throughput phenotypic profiling

- BiologybioRxiv
- 2023

A novel Latent Eigenvalue Analysis (LEA) framework is proposed and applied to high-throughput phenotypic profiling with single-cell and single-organelle granularity and achieves a robust quantification of phenotypesic changes introduced by drug treatment.

## References

SHOWING 1-10 OF 39 REFERENCES

### Causal inference by using invariant prediction: identification and confidence intervals

- Computer Science, Mathematics
- 2015

This work proposes to exploit invariance of a prediction under a causal model for causal inference: given different experimental settings (e.g. various interventions) the authors collect all models that do show invariance in their predictive accuracy across settings and interventions, and yields valid confidence intervals for the causal relationships in quite general scenarios.

### Interventional Experiment Design for Causal Structure Learning

- Computer ScienceArXiv
- 2019

This work considers a setup in which given a budget $k, the problem of finding the best intervention target set as an optimization problem which aims to maximize the number of edges whose directions are identified due to the performed interventions, and proposes an efficient exact calculator as well as two estimators for this task.

### Invariant Causal Prediction for Sequential Data

- MathematicsJournal of the American Statistical Association
- 2018

ABSTRACT We investigate the problem of inferring the causal predictors of a response Y from a set of d explanatory variables (X1, …, Xd). Classical ordinary least-square regression includes all…

### Invariant Causal Prediction for Nonlinear Models

- Computer ScienceJournal of Causal Inference
- 2018

This work presents and evaluates an array of methods for nonlinear and nonparametric versions of ICP for learning the causal parents of given target variables and finds that an approach which first fits a nonlinear model with data pooled over all environments and then tests for differences between the residual distributions across environments is quite robust across a large variety of simulation settings.

### Two optimal strategies for active learning of causal models from interventional data

- Computer ScienceInt. J. Approx. Reason.
- 2014

### Active Learning of Causal Networks with Intervention Experiments and Optimal Designs

- Computer Science
- 2008

It is shown theoretically that structural learning can be done locally in subgraphs of chain components without need of checking illegal v-structures and cycles in the whole network and that a Markov equivalence subclass obtained after each intervention can still be depicted as a chain graph.

### ABCD-Strategy: Budgeted Experimental Design for Targeted Causal Structure Discovery

- Computer ScienceAISTATS
- 2019

This work proposes a general framework and theory based on optimal Bayesian experimental design to select experiments for targeted causal discovery and provides a tractable implementation with provable guarantees on both approximation and optimization quality based on submodularity.

### Anchor regression: Heterogeneous data meet causality

- MathematicsJournal of the Royal Statistical Society: Series B (Statistical Methodology)
- 2021

We consider the problem of predicting a response variable from a set of covariates on a data set that differs in distribution from the training data. Causal parameters are optimal in terms of…

### Preventing Failures Due to Dataset Shift: Learning Predictive Models That Transport

- Computer ScienceAISTATS
- 2019

It is proved that the surgery estimator finds stable relationships in strictly more scenarios than previous approaches which only consider conditional relationships, and performs competitively against entirely data-driven approaches.

### Stabilizing variable selection and regression

- Mathematics, Computer ScienceThe Annals of Applied Statistics
- 2021

St stabilized regression is introduced which explicitly enforces stability and thus improves generalization performance to previously unseen environments and it is proved that this set of predictors is optimal in the sense that a regression based on these predictors minimizes the mean squared prediction error given that the resulting regression generalizes to unseen new environments.