2021

Learning Neural Causal Models with Active Interventions

Scherrer, Nino, Bilaniuk, Olexa, Annadani, Yashas et al.

Understand

Discovering causal structures from data is a challenging inference problem of fundamental importance in all areas of science.

The appealing properties of neural networks have recently led to a surge of interest in differentiable neural network-based methods for learning causal structures from data.
So far, differentiable causal discovery has focused on static datasets of observational or fixed interventional origin.
In this work, we introduce an active intervention targeting (AIT) method which enables a quick identification of the underlying causal structure of the data-generating process.

Built on

Individual Choice Behavior: A Theoretical analysis
Luce, R. D · 1959
Earlier work this paper cites.
Optimal design: Experiments for discriminating between several models
Atkinson, A. C. and Fedorov, V. V · 1975
Earlier work this paper cites.
The analysis of permutations
Plackett, R. L · 1975
Earlier work this paper cites.
Local computations with probabilities on graphical structures and their application to expert systems
Lauritzen, S. L. and Spiegelhalter, D. J · 1988
Earlier work this paper cites.
Learning bayesian networks: The combination of knowledge and statistical data
Heckerman, D., Geiger, D., and Chickering, D. M · 1995
Earlier work this paper cites.
Using bayesian networks to analyze expression data
Friedman, N., Linial, M., Nachman, I., and Pe’er, D · 2000
Earlier work this paper cites.
Marginal structural models and causal inference in epidemiology, 2000
Robins, J. M., Hernan, M. A., and Brumback, B · 2000
Earlier work this paper cites.
Causation, prediction, and search
Spirtes, P., Glymour, C. N., Scheines, R., Heckerman, D., Meek, C., Cooper, G., and Richardson, T · 2000
Earlier work this paper cites.
Active learning of causal bayes net structure
Murphy, K. P · 2001
Earlier work this paper cites.
Active learning for structure in bayesian networks
Tong, S. and Koller, D · 2001
Earlier work this paper cites.
Optimal structure identification with greedy search
Chickering, D. M · 2002
Earlier work this paper cites.
Searching for bayesian network structures in the space of restricted acyclic partially directed graphs
Acid, S. and de Campos, L. M · 2003
Earlier work this paper cites.
Causal protein-signaling networks derived from multiparameter single-cell data
Sachs, K., Perez, O., Pe’er, D., Lauffenburger, D. A., and Nolan, G. P · 2005
Earlier work this paper cites.
Interventions and causal inference
Eberhardt, F. and Scheines, R · 2007
Earlier work this paper cites.
Active learning of causal networks with intervention experiments and optimal designs
He, Y.-B. and Geng, Z · 2008
Earlier work this paper cites.
Bayesian artificial intelligence
Korb, K. B. and Nicholson, A. E · 2010
Earlier work this paper cites.
Identifiability of causal graphs using functional models
Peters, J., Mooij, J. M., Janzing, D., and Schölkopf, B · 2011
Earlier work this paper cites.
Almost optimal intervention sets for causal discovery
Original
Eberhardt, F · 2012
Earlier work this paper cites.
On the number of experiments sufficient and in the worst case necessary to identify all causal relations among n variables
Original
Eberhardt, F., Glymour, C., and Scheines, R · 2012
Earlier work this paper cites.
Characterization and greedy learning of interventional markov equivalence classes of directed acyclic graphs
Hauser, A. and Bühlmann, P · 2012
Earlier work this paper cites.

Similar

Experiment selection for causal discovery
Hyttinen, A., Eberhardt, F., and Hoyer, P. O · 2013
Cited alongside, same era.
An interactive approach for bayesian network learning using domain/expert knowledge
Masegosa, A. R. and Moral, S · 2013
Cited alongside, same era.
Two optimal strategies for active learning of causal models from interventional data
Hauser, A. and Bühlmann, P · 2014
Cited alongside, same era.
Variational inference with normalizing flows
Original
Rezende, D. J. and Mohamed, S · 2015
Cited alongside, same era.
Learning causal graphs with small interventions
Shanmugam, K., Kocaoglu, M., Dimakis, A. G., and Vishwanath, S · 2015
Cited alongside, same era.
Reconstructing causal biological networks through active learning
Cho, H., Berger, B., and Peng, J · 2016
Cited alongside, same era.
Inferring causal molecular networks: empirical assessment through a community-based effort
Hill, S. M., Heiser, L. M., Cokelaer, T., Unger, M., Nesser, N. K., Carlin, D. E., Zhang, Y., Sokolov, A., Paull, E. O., Wong, C. K., et al · 2016
Cited alongside, same era.
Causal inference by using invariant prediction: identification and confidence intervals
Peters, J., Bühlmann, P., and Meinshausen, N · 2016
Cited alongside, same era.
Causality and causal inference in epidemiology: the need for a pluralistic approach
Vandenbroucke, J. P., Broadbent, A., and Pearce, N · 2016
Cited alongside, same era.
A bayesian active learning experimental design for inferring signaling networks
Ness, R. O., Sachs, K., Mallick, P., and Vitek, O · 2017
Cited alongside, same era.
Elements of causal inference: foundations and learning algorithms
Peters, J., Janzing, D., and Schölkopf, B · 2017
Cited alongside, same era.
Budgeted experiment design for causal structure learning
Ghassami, A., Salehkaleybar, S., Kiyavash, N., and Bareinboim, E · 2018
Cited alongside, same era.

Then

Sample efficient active learning of causal trees
Greenewald, K., Katz, D., Shanmugam, K., Magliacane, S., Kocaoglu, M., Adsera, E. B., and Bresler, G · 2019
Later among the works it cites.
Causal discovery toolbox: Uncover causal relationships in python
Original
Kalainathan, D. and Goudet, O · 2019
Later among the works it cites.
Learning neural causal models from unknown interventions
Original
Ke, N. R., Bilaniuk, O., Goyal, A., Bauer, S., Larochelle, H., Schölkopf, B., Mozer, M. C., Pal, C., and Bengio, Y · 2019
Later among the works it cites.
A graph autoencoder approach to causal structure learning
Original
Ng, I., Zhu, S., Chen, Z., and Fang, Z · 2019
Later among the works it cites.
Dag-gnn: Dag structure learning with graph neural networks
Yu, Y., Chen, J., Gao, T., and Yu, M · 2019
Later among the works it cites.
Differentiable causal discovery from interventional data
Brouillard, P., Lachapelle, S., Lacoste, A., Lacoste-Julien, S., and Drouin, A · 2020
Later among the works it cites.
Active invariant causal prediction: Experiment selection through stability
Original
Gamella, J. L. and Heinze-Deml, C · 2020
Later among the works it cites.
Gradient-based neural dag learning
Lachapelle, S., Brouillard, P., Deleu, T., and Lacoste-Julien, S · 2020
Later among the works it cites.
Joint causal inference from multiple contexts
Mooij, J. M., Magliacane, S., and Claassen, T · 2020
Later among the works it cites.
Active structure learning of causal dags via directed clique tree
Original
Squires, C., Magliacane, S., Greenewald, K., Katz, D., Kocaoglu, M., and Shanmugam, K · 2020
Later among the works it cites.
Learning sparse nonparametric dags
Zheng, X., Dan, C., Aragam, B., Ravikumar, P., and Xing, E · 2020
Later among the works it cites.
Causal discovery with reinforcement learning
Zhu, S., Ng, I., and Chen, Z · 2020
Later among the works it cites.
Variational causal networks: Approximate bayesian inference over causal structures
Original
Annadani, Y., Rothfuss, J., Lacoste, A., Scherrer, N., Goyal, A., Bengio, Y., and Bauer, S · 2021
Closest in time.
Bcd nets: Scalable variational approaches for bayesian causal discovery
Cundy, C., Grover, A., and Ermon, S · 2021
Closest in time.
Efficient neural causal discovery without acyclicity constraints
Original
Lippe, P., Cohen, T., and Gavves, E · 2021
Closest in time.
Dibs: Differentiable bayesian structure learning
Original
Lorch, L., Rothfuss, J., Schölkopf, B., and Krause, A · 2021
Closest in time.
Toward causal representation learning
Schölkopf, B., Locatello, F., Bauer, S., Ke, N. R., Kalchbrenner, N., Goyal, A., and Bengio, Y · 2021
Closest in time.
D’ya like dags? a survey on structure learning and causal discovery
Original
Vowels, M. J., Camgoz, N. C., and Bowden, R · 2021
Closest in time.
The causal-neural connection: Expressiveness, learnability, and inference
Xia, K., Lee, K.-Z., Bengio, Y., and Bareinboim, E · 2021
Closest in time.
Bayesian optimal experimental design for inferring causal structure
Original
Zemplenyi, M. and Miller, J. W · 2021
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

Reading the bibliography…