Recent advances in probabilistic generative modeling have motivated learning Structural Causal Models (SCM) from observational datasets using deep conditional generative models, also known as Deep Structural Causal Models (DSCM). If successful, DSCMs can be utilized for causal estimation tasks, e.g., for answering counterfactual queries. In this work, we warn practitioners about non-identifiability of counterfactual inference from observational data, even in the absence of unobserved confounding and assuming known causal structure. We prove counterfactual identifiability of monotonic generation mechanisms with single dimensional exogenous variables. For general generation mechanisms with multi-dimensional exogenous variables, we provide an impossibility result for counterfactual identifiability, motivating the need for parametric assumptions. As a practical approach, we propose a method for estimating worst-case errors of learned DSCMs' counterfactual predictions. The size of this error can be an essential metric for deciding whether or not DSCMs are a viable approach for counterfactual inference in a specific problem setting. In evaluation, our method confirms negligible counterfactual errors for an identifiable SCM from prior work, and also provides informative error bounds on counterfactual errors for a non-identifiable synthetic SCM.
By taking into account the nonlinear effect of the cause, the inner noise effect, and the measurement distortion effect in the observed variables, the post-nonlinear (PNL) causal model has demonstrated its excellent performance in distinguishing the cause from effect. However, its identifiability has not been properly addressed, and how to apply it in the case of more than two variables is also a problem. In this paper, we conduct a systematic investigation on its identifiability in the two-variable case. We show that this model is identifiable in most cases; by enumerating all possible situat…
University of Copenhagen（哥本哈根大学）；Technical University of Denmark（丹麦技术大学）；University of Amsterdam（阿姆斯特丹大学）
This PhD thesis contains several contributions to the field of statistical causal modeling. Statistical causal models are statistical models embedded with causal assumptions that allow for the inference and reasoning about the behavior of stochastic systems affected by external manipulation (interventions). This thesis contributes to the research areas concerning the estimation of causal effects, causal structure learning, and distributionally robust (out-of-distribution generalizing) prediction methods. We present novel and consistent linear and non-linear causal effects estimators in instrum…
Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所）；Radboud University Nijmegen（拉德堡德大学奈梅亨分校）
This work addresses the following question: Under what assumptions on the data generating process can one infer the causal graph from the joint distribution? The approach taken by conditional independence-based causal discovery methods is based on two assumptions: the Markov condition and faithfulness. It has been shown that under these assumptions the causal graph can be identified up to Markov equivalence (some arrows remain undirected) using methods like the PC algorithm. In this work we propose an alternative by defining Identifiable Functional Model Classes (IFMOCs). As our main theorem w…
Multi-View Causal Discovery without Non-Gaussianity: Identifiability and Algorithms
Inria（法国国家信息与自动化技术研究院）；CEA（法国国家科学研究中心）；University of Paris-Saclay（巴黎-萨克勒大学）；University of Carnegie Mellon（卡内基梅隆大学）；Apple（苹果公司）；University of Helsinki（赫尔辛基大学）
We formulate a general framework for building structural causal models (SCMs) with deep learning components. The proposed approach employs normalising flows and variational inference to enable tractable inference of exogenous noise variables - a crucial step for counterfactual inference that is missing from existing deep causal learning methods. Our framework is validated on a synthetic dataset built on MNIST as well as on a real-world medical dataset of brain MRI scans. Our experimental results indicate that we can successfully train deep SCMs that are capable of all three levels of Pearl's l…
Efficient Symbolic Computations for Identifying Causal Effects
Max Planck Institute for Mathematics in the Sciences（马克斯·普朗克数学研究院）；Binghamton University（宾夕法尼亚州立大学布林茅尔分校）；École Polytechnique Fédérale de Lausanne（洛桑联邦理工学院）
Counterfactual explanations (CEs) are methods for generating an alternative scenario that produces a different desirable outcome. For example, if a student is predicted to fail a course, then counterfactual explanations can provide the student with alternate ways so that they would be predicted to pass. The applications are many. However, CEs are currently generated from machine learning models that do not necessarily take into account the true causal structure in the data. By doing this, bias can be introduced into the CE quantities. I propose in this study to test the CEs using Judea Pearl's…
University College London（伦敦大学学院）；The Alan Turing Institute（艾伦·图灵研究所）
In this paper, we address the challenge of performing counterfactual inference with observational data via Bayesian nonparametric regression adjustment, with a focus on high-dimensional settings featuring multiple actions and multiple correlated outcomes. We present a general class of counterfactual multi-task deep kernels models that estimate causal effects and learn policies proficiently thanks to their sample efficiency gains, while scaling well with high dimensions. In the first part of the work, we rely on Structural Causal Models (SCM) to formally introduce the setup and the problem of i…
Massachusetts Institute of Technology（麻省理工学院）；CSAIL（计算机科学与人工智能实验室）
We introduce an off-policy evaluation procedure for highlighting episodes where applying a reinforcement learned (RL) policy is likely to have produced a substantially different outcome than the observed policy. In particular, we introduce a class of structural causal models (SCMs) for generating counterfactual trajectories in finite partially observable Markov Decision Processes (POMDPs). We see this as a useful procedure for off-policy "debugging" in high-risk settings (e.g., healthcare); by decomposing the expected difference in reward between the RL and observed policy into specific episod…
The notion of causality is used in many situations dealing with uncertainty. We consider the problem whether causality can be identified given data set generated by discrete random variables rather than continuous ones. In particular, for non-binary data, thus far it was only known that causality can be identified except rare cases. In this paper, we present necessary and sufficient condition for an integer modular acyclic additive noise (IMAN) of two variables. In addition, we relate bivariate and multivariate causal identifiability in a more explicit manner, and develop a practical algorithm…
College of Information and Computer Sciences, University of Massachusetts, Amherst（马萨诸塞大学阿默斯特分校信息与计算机科学学院）；Basis Research Institute（贝西斯研究院）；Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology（麻省理工学院脑与认知科学系）
A growing family of approaches to causal inference rely on Bayesian formulations of assumptions that go beyond causal graph structure. For example, Bayesian approaches have been developed for analyzing instrumental variable designs, regression discontinuity designs, and within-subjects designs. This paper introduces simulation-based identifiability (SBI), a procedure for testing the identifiability of queries in Bayesian causal inference approaches that are implemented as probabilistic programs. SBI complements analytical approaches to identifiability, leveraging a particle-based optimization …
University of Maryland, College Park（马里兰大学帕克分校）；Indian Institute of Science, Bangalore（班加罗尔印度科学研究所）；Microsoft Research, India（微软研究院（印度））
In this work, we consider the problem of robust parameter estimation from observational data in the context of linear structural equation models (LSEMs). LSEMs are a popular and well-studied class of models for inferring causality in the natural and social sciences. One of the main problems related to LSEMs is to recover the model parameters from the observational data. Under various conditions on LSEMs and the model parameters the prior work provides efficient algorithms to recover the parameters. However, these results are often about generic identifiability. In practice, generic identifiabi…
We consider basic conceptual questions concerning the relationship between statistical estimation and causal inference. Firstly, we show how to translate causal inference problems into an abstract statistical formalism without requiring any structure beyond an arbitrarily-indexed family of probability models. The formalism is simple but can incorporate a variety of causal modelling frameworks, including 'structural causal models', but also models expressed in terms of, e.g., differential equations. We focus primarily on the structural/graphical causal modelling literature, however. Secondly, w…
This manuscript contributes a general and practical framework for casting a Markov process model of a system at equilibrium as a structural causal model, and carrying out counterfactual inference. Markov processes mathematically describe the mechanisms in the system, and predict the system's equilibrium behavior upon intervention, but do not support counterfactual inference. In contrast, structural causal models support counterfactual inference, but do not identify the mechanisms. This manuscript leverages the benefits of both approaches. We define the structural causal models in terms of the …
Carnegie Mellon University（卡内基梅隆大学）；University of Pittsburgh（匹兹堡大学）
Measurement error in the observed values of the variables can greatly change the output of various causal discovery methods. This problem has received much attention in multiple fields, but it is not clear to what extent the causal model for the measurement-error-free variables can be identified in the presence of measurement error with unknown variance. In this paper, we study precise sufficient identifiability conditions for the measurement-error-free causal model and show what information of the causal model can be recovered from observed data. In particular, we present two different sets o…
Evaluating hypothetical statements about how the world would be had a different course of action been taken is arguably one key capability expected from modern AI systems. Counterfactual reasoning underpins discussions in fairness, the determination of blame and responsibility, credit assignment, and regret. In this paper, we study the evaluation of counterfactual statements through neural models. Specifically, we tackle two causal problems required to make such evaluations, i.e., counterfactual identification and estimation from an arbitrary combination of observational and experimental data.…
Identifiability Guarantees for Causal Disentanglement from Soft Interventions
Broad Institute of MIT and Harvard（麻省理工学院与哈佛大学布罗德研究所）；Laboratory for Information and Decision Systems, MIT（麻省理工学院信息与决策系统实验室）；MIT-IBM Watson AI Lab（MIT-IBM沃森人工智能实验室）；IBM Research（IBM研究院）；Google Research（谷歌研究院）
Technion（以色列理工学院）；Google Research（谷歌研究院）；University of Toronto（多伦多大学）；Vector Institute（向量研究所）
To perform counterfactual reasoning in Structural Causal Models (SCMs), one needs to know the causal mechanisms, which provide factorizations of conditional distributions into noise sources and deterministic functions mapping realizations of noise to samples. Unfortunately, the causal mechanism is not uniquely identified by data that can be gathered by observing and interacting with the world, so there remains the question of how to choose causal mechanisms. In recent work, Oberst & Sontag (2019) propose Gumbel-max SCMs, which use Gumbel-max reparameterizations as the causal mechanism due to a…
Microsoft Research（微软研究院）；University of Colorado Boulder（科罗拉多大学博尔德分校）
To construct interpretable explanations that are consistent with the original ML model, counterfactual examples---showing how the model's output changes with small perturbations to the input---have been proposed. This paper extends the work in counterfactual explanations by addressing the challenge of feasibility of such examples. For explanations of ML models in critical domains such as healthcare and finance, counterfactual examples are useful for an end-user only to the extent that perturbation of feature inputs is feasible in the real world. We formulate the problem of feasibility as prese…
Imperial College London（帝国理工学院）；Spotify（Spotify公司）；FAU Erlangen-Nuremberg（埃尔朗根-纽伦堡大学）；University College London（伦敦大学学院）
Counterfactual inference is a powerful tool, capable of solving challenging problems in high-profile sectors. To perform counterfactual inference, one requires knowledge of the underlying causal mechanisms. However, causal mechanisms cannot be uniquely determined from observations and interventions alone. This raises the question of how to choose the causal mechanisms so that resulting counterfactual inference is trustworthy in a given domain. This question has been addressed in causal models with binary variables, but the case of categorical variables remains unanswered. We address this chall…
We consider the task of counterfactual estimation from observational imaging data given a known causal structure. In particular, quantifying the causal effect of interventions for high-dimensional data with neural networks remains an open challenge. Herein we propose Diff-SCM, a deep structural causal model that builds on recent advances of generative energy-based models. In our setting, inference is performed by iteratively sampling gradients of the marginal and conditional distributions entailed by the causal model. Counterfactual estimation is achieved by firstly inferring latent variables …
Microsoft Research（微软研究院）；Indian Institute of Technology Hyderabad（印度理工学院海得拉巴分校）
Counterfactual examples for an input -- perturbations that change specific features but not others -- have been shown to be useful for evaluating bias of machine learning models, e.g., against specific demographic groups. However, generating counterfactual examples for images is non-trivial due to the underlying causal structure on the various features of an image. To be meaningful, generated perturbations need to satisfy constraints implied by the causal model. We present a method for generating counterfactuals by incorporating a structural causal model (SCM) in an improved variant of Adversa…
MIT（麻省理工学院）；MIT-IBM Watson AI Lab（麻省理工学院-IBM沃森人工智能实验室）；IBM Research（IBM研究院）
Entropic causal inference is a framework for inferring the causal direction between two categorical variables from observational data. The central assumption is that the amount of unobserved randomness in the system is not too large. This unobserved randomness is measured by the entropy of the exogenous variable in the underlying structural causal model, which governs the causal relation between the observed variables. Kocaoglu et al. conjectured that the causal direction is identifiable when the entropy of the exogenous variable is not too large. In this paper, we prove a variant of their con…
University of Amsterdam（阿姆斯特丹大学）；MIT-IBM Watson AI Lab（麻省理工学院-IBM沃森人工智能实验室）；Qualcomm AI Research（高通人工智能研究院）
Understanding the latent causal factors of a dynamical system from visual observations is considered a crucial step towards agents reasoning in complex environments. In this paper, we propose CITRIS, a variational autoencoder framework that learns causal representations from temporal sequences of images in which underlying causal factors have possibly been intervened upon. In contrast to the recent literature, CITRIS exploits temporality and observing intervention targets to identify scalar and multidimensional causal factors, such as 3D rotation angles. Furthermore, by introducing a normalizi…
Identifying the effects of causes and causes of effects is vital in virtually every scientific field. Often, however, the needed probabilities may not be fully identifiable from the data sources available. This paper shows how partial identifiability is still possible for several probabilities of causation. We term this epsilon-identifiability and demonstrate its usefulness in cases where the behavior of certain subpopulations can be restricted to within some narrow bounds. In particular, we show how unidentifiable causal effects and counterfactual probabilities can be narrowly bounded when su…
We address the problem of estimating the effect of intervening on a set of variables X from experiments on a different set, Z, that is more accessible to manipulation. This problem, which we call z-identifiability, reduces to ordinary identifiability when Z = empty and, like the latter, can be given syntactic characterization using the do-calculus [Pearl, 1995; 2000]. We provide a graphical necessary and sufficient condition for z-identifiability for arbitrary sets X,Z, and Y (the outcomes). We further develop a complete algorithm for computing the causal effect of X on Y using information pro…
University of Technology Sydney（悉尼科技大学）；Curtin University（科廷大学）
Counterfactual fairness alleviates the discrimination between the model prediction toward an individual in the actual world (observational data) and that in counterfactual world (i.e., what if the individual belongs to other sensitive groups). The existing studies need to pre-define the structural causal model that captures the correlations among variables for counterfactual inference; however, the underlying causal model is usually unknown and difficult to be validated in real-world scenarios. Moreover, the misspecification of the causal model potentially leads to poor performance in model pr…
Korea Advanced Institute of Science and Technology (KAIST)（韩国科学技术院）
The problem of fair classification can be mollified if we develop a method to remove the embedded sensitive information from the classification features. This line of separating the sensitive information is developed through the causal inference, and the causal inference enables the counterfactual generations to contrast the what-if case of the opposite sensitive attribute. Along with this separation with the causality, a frequent assumption in the deep latent causal model defines a single latent variable to absorb the entire exogenous uncertainty of the causal graph. However, we claim that su…
Identifiable Latent Polynomial Causal Models Through the Lens of Change
Australian Institute for Machine Learning, The University of Adelaide（阿德莱德大学澳大利亚机器学习研究所）；School of Computer Science and Engineering, The University of New South Wales（新南威尔士大学计算机科学与工程学院）；School of Mathematics and Statistics, The University of Melbourne（墨尔本大学数学与统计学院）；Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）；Halicioğlu Data Science Institute (HDSI), University of California San Diego（加州大学圣地亚哥分校哈利西奥格鲁数据科学研究所）；Carnegie Mellon University（卡内基梅隆大学）
Variational autoencoders (VAEs) and other generative methods have garnered growing interest not just for their generative properties but also for the ability to dis-entangle a low-dimensional latent variable space. However, few existing generative models take causality into account. We propose a new decoder based framework named the Causal Counterfactual Generative Model (CCGM), which includes a partially trainable causal layer in which a part of a causal model can be learned without significantly impacting reconstruction fidelity. By learning the causal relationships between image semantic la…
From What Ifs to Insights: Counterfactuals in Causal Inference vs. Explainable AI
Institute of Service Science, National Tsing Hua University（服务科学研究所，国立清华大学）；Faculty of Business and Economics, University of Antwerp（商学院与经济学院，安特卫普大学）；Copenhagen Business School（哥本哈根商学院）
CausalGaze: Unveiling Hallucinations via Counterfactual Graph Intervention in Large Language Models
College of Electronic Engineering, National University of Defense Technology（国防科技大学电子工程学院）；Anhui Province Key Laboratory of Cyberspace Security Situation Awareness and Evaluation（安徽省网络空间安全态势感知与评估重点实验室）；Institute of Computer Application, China Academy of Engineering Physics（中国工程物理研究院计算机应用研究所）
On the Parameter Identifiability of Partially Observed Linear Causal Models
Carnegie Mellon University（卡内基梅隆大学）；University of California San Diego（加州大学圣迭戈分校）；Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）；KDDI Research（KDDI研究所）
