Causal ABMs: Learning Plausible Causal Models using Agent-based Modeling

Konstantina Valogianni, Balaji Padmanabhan
Proceedings of The KDD'22 Workshop on Causal Discovery, PMLR 185:3-29, 2022.

Abstract

We present Causal ABM, a methodology to derive causal structures describing complex underlying behavioral phenomena. Agent-based models (ABMs) have powerful advantages for causal modeling that have not been explored sufficiently. Unlike traditional causal estimation approaches which often result in “one best” causal structure that is learned, two properties of ABMs - equifinality (the ability of different sets of conditions or model representations to yield the same outcome) and mutlifinality (the same ABM might yield different outcomes) - can be exploited to learn multiple diverse “plausible causal models” from data. Using an illustrative example of news sharing on social networks we show how this idea can be applied to learn such causal sets. We also show how genetic algorithms can be used as a estimation technique to learn multiple plausible causal models from data due to their parallel search structure. However, significant computational challenges remain before this can be generally applied, and we, therefore, highlight specific key issues that need to be addressed in future work.

Cite this Paper


BibTeX
@InProceedings{pmlr-v185-valogianni22a, title = {Causal ABMs: Learning Plausible Causal Models using Agent-based Modeling }, author = {Valogianni, Konstantina and Padmanabhan, Balaji}, booktitle = {Proceedings of The KDD'22 Workshop on Causal Discovery}, pages = {3--29}, year = {2022}, editor = {Le, Thuc Duy and Liu, Lin and Kıcıman, Emre and Triantafyllou, Sofia and Liu, Huan}, volume = {185}, series = {Proceedings of Machine Learning Research}, month = {15 Aug}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v185/valogianni22a/valogianni22a.pdf}, url = {https://proceedings.mlr.press/v185/valogianni22a.html}, abstract = {We present Causal ABM, a methodology to derive causal structures describing complex underlying behavioral phenomena. Agent-based models (ABMs) have powerful advantages for causal modeling that have not been explored sufficiently. Unlike traditional causal estimation approaches which often result in “one best” causal structure that is learned, two properties of ABMs - equifinality (the ability of different sets of conditions or model representations to yield the same outcome) and mutlifinality (the same ABM might yield different outcomes) - can be exploited to learn multiple diverse “plausible causal models” from data. Using an illustrative example of news sharing on social networks we show how this idea can be applied to learn such causal sets. We also show how genetic algorithms can be used as a estimation technique to learn multiple plausible causal models from data due to their parallel search structure. However, significant computational challenges remain before this can be generally applied, and we, therefore, highlight specific key issues that need to be addressed in future work.} }
Endnote
%0 Conference Paper %T Causal ABMs: Learning Plausible Causal Models using Agent-based Modeling %A Konstantina Valogianni %A Balaji Padmanabhan %B Proceedings of The KDD'22 Workshop on Causal Discovery %C Proceedings of Machine Learning Research %D 2022 %E Thuc Duy Le %E Lin Liu %E Emre Kıcıman %E Sofia Triantafyllou %E Huan Liu %F pmlr-v185-valogianni22a %I PMLR %P 3--29 %U https://proceedings.mlr.press/v185/valogianni22a.html %V 185 %X We present Causal ABM, a methodology to derive causal structures describing complex underlying behavioral phenomena. Agent-based models (ABMs) have powerful advantages for causal modeling that have not been explored sufficiently. Unlike traditional causal estimation approaches which often result in “one best” causal structure that is learned, two properties of ABMs - equifinality (the ability of different sets of conditions or model representations to yield the same outcome) and mutlifinality (the same ABM might yield different outcomes) - can be exploited to learn multiple diverse “plausible causal models” from data. Using an illustrative example of news sharing on social networks we show how this idea can be applied to learn such causal sets. We also show how genetic algorithms can be used as a estimation technique to learn multiple plausible causal models from data due to their parallel search structure. However, significant computational challenges remain before this can be generally applied, and we, therefore, highlight specific key issues that need to be addressed in future work.
APA
Valogianni, K. & Padmanabhan, B.. (2022). Causal ABMs: Learning Plausible Causal Models using Agent-based Modeling . Proceedings of The KDD'22 Workshop on Causal Discovery, in Proceedings of Machine Learning Research 185:3-29 Available from https://proceedings.mlr.press/v185/valogianni22a.html.

Related Material