Guardado en:
Detalles Bibliográficos
Autores principales: Fujii, Keisuke, Takeuchi, Koh, Kuribayashi, Atsushi, Takeishi, Naoya, Kawahara, Yoshinobu, Takeda, Kazuya
Formato: Preprint
Publicado: 2022
Materias:
Acceso en línea:https://arxiv.org/abs/2206.01900
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929246102880256
author Fujii, Keisuke
Takeuchi, Koh
Kuribayashi, Atsushi
Takeishi, Naoya
Kawahara, Yoshinobu
Takeda, Kazuya
author_facet Fujii, Keisuke
Takeuchi, Koh
Kuribayashi, Atsushi
Takeishi, Naoya
Kawahara, Yoshinobu
Takeda, Kazuya
contents Evaluation of intervention in a multiagent system, e.g., when humans should intervene in autonomous driving systems and when a player should pass to teammates for a good shot, is challenging in various engineering and scientific fields. Estimating the individual treatment effect (ITE) using counterfactual long-term prediction is practical to evaluate such interventions. However, most of the conventional frameworks did not consider the time-varying complex structure of multiagent relationships and covariate counterfactual prediction. This may lead to erroneous assessments of ITE and difficulty in interpretation. Here we propose an interpretable, counterfactual recurrent network in multiagent systems to estimate the effect of the intervention. Our model leverages graph variational recurrent neural networks and theory-based computation with domain knowledge for the ITE estimation framework based on long-term prediction of multiagent covariates and outcomes, which can confirm the circumstances under which the intervention is effective. On simulated models of an automated vehicle and biological agents with time-varying confounders, we show that our methods achieved lower estimation errors in counterfactual covariates and the most effective treatment timing than the baselines. Furthermore, using real basketball data, our methods performed realistic counterfactual predictions and evaluated the counterfactual passes in shot scenarios.
format Preprint
id arxiv_https___arxiv_org_abs_2206_01900
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Estimating counterfactual treatment outcomes over time in complex multiagent scenarios
Fujii, Keisuke
Takeuchi, Koh
Kuribayashi, Atsushi
Takeishi, Naoya
Kawahara, Yoshinobu
Takeda, Kazuya
Artificial Intelligence
Machine Learning
Multiagent Systems
Methodology
Evaluation of intervention in a multiagent system, e.g., when humans should intervene in autonomous driving systems and when a player should pass to teammates for a good shot, is challenging in various engineering and scientific fields. Estimating the individual treatment effect (ITE) using counterfactual long-term prediction is practical to evaluate such interventions. However, most of the conventional frameworks did not consider the time-varying complex structure of multiagent relationships and covariate counterfactual prediction. This may lead to erroneous assessments of ITE and difficulty in interpretation. Here we propose an interpretable, counterfactual recurrent network in multiagent systems to estimate the effect of the intervention. Our model leverages graph variational recurrent neural networks and theory-based computation with domain knowledge for the ITE estimation framework based on long-term prediction of multiagent covariates and outcomes, which can confirm the circumstances under which the intervention is effective. On simulated models of an automated vehicle and biological agents with time-varying confounders, we show that our methods achieved lower estimation errors in counterfactual covariates and the most effective treatment timing than the baselines. Furthermore, using real basketball data, our methods performed realistic counterfactual predictions and evaluated the counterfactual passes in shot scenarios.
title Estimating counterfactual treatment outcomes over time in complex multiagent scenarios
topic Artificial Intelligence
Machine Learning
Multiagent Systems
Methodology
url https://arxiv.org/abs/2206.01900