884 Hits in 2.7 sec

Causal Graphical Models with Latent Variables: Learning and Inference [chapter]

Stijn Meganck, Philippe Leray, Bernard Manderick
2007 Lecture Notes in Computer Science  
We will do this by proposing an alternative representation for semi-Markovian causal models.  ...  Previously an algorithm has been constructed that by combining elements from both techniques allows to learn a semi-Markovian causal models from a mixture of observational and experimental data.  ...  semi-Markovian causal models.  ... 
doi:10.1007/978-3-540-75256-1_4 fatcat:jritqd44njevjh3a64au433awy

Identifiability of Causal-based Fairness Notions: A State of the Art [article]

Karima Makhlouf, Sami Zhioua, Catuscia Palamidessi
2022 arXiv   pre-print
The results are illustrated using a large number of examples and causal graphs.  ...  The paper would be of particular interest to fairness researchers, practitioners, and policy makers who are considering the use of causality-based fairness notions as it summarizes and illustrates the  ...  Identifiable semi-Markovian models Causal effects are not always identifiable in semi-Markovian models. This subsection focuses on causal models where the causal effect of on is identifiable.  ... 
arXiv:2203.05900v1 fatcat:jks3yjbczbfmvkiv2vwzewfczu

Survey on Causal-based Machine Learning Fairness Notions [article]

Karima Makhlouf, Sami Zhioua, Catuscia Palamidessi
2022 arXiv   pre-print
Addressing the problem of fairness is crucial to safely use machine learning algorithms to support decisions with a critical impact on people's lives such as job hiring, child maltreatment, disease diagnosis  ...  The most recent fairness notions, however, are causal-based and reflect the now widely accepted idea that using causality is necessary to appropriately address the problem of fairness.  ...  (a)), semi-Markovian model (Figures 2(b)) and semi-Markovian model after intervening on (Figure 2(c)). Figure 2 : 2 Figure 2: Markovian and semi-Markovian causal models.  ... 
arXiv:2010.09553v6 fatcat:f4cqm2hjlfhdbewixcsd27dhxy

Variational Latent-State GPT for Semi-supervised Task-Oriented Dialog Systems [article]

Hong Liu, Yucheng Cai, Zhenru Lin, Zhijian Ou, Yi Huang, Junlan Feng
2022 arXiv   pre-print
The inference model in VLS-GPT is non-Markovian due to the use of the Transformer architecture.  ...  Semi-supervised TOD experiments are conducted on two benchmark multi-domain datasets of different languages - MultiWOZ2.1 and CrossWOZ.  ...  than previous variational learning works for sequential latent variable models [13] , [27] , which use turn-level first-order Markovian.  ... 
arXiv:2109.04314v2 fatcat:j6toboql2bezjl67zibl5bcaey

Anti-discrimination learning: a causal modeling-based framework

Lu Zhang, Xintao Wu
2017 International Journal of Data Science and Analytics  
In this paper, we introduce a causal modeling-based framework for anti-discrimination learning.  ...  The aim of this paper is to deepen the understanding of discrimination in data mining from the causal modeling perspective, and suggest several potential future research directions.  ...  The situation in the semi-Markovian model is much more complicated than that in the Markovian model, both in the causal graph learning and causal effect inference.  ... 
doi:10.1007/s41060-017-0058-x dblp:journals/ijdsa/ZhangW17 fatcat:lal6fscwongc7d5gz7hxab4774

Inequality Constraints in Causal Models with Hidden Variables [article]

Changsung Kang, Jin Tian
2012 arXiv   pre-print
The results have applications in testing causal models with observational or experimental data.  ...  We derive bounds on causal effects that are not directly measured in randomized experiments. We derive instrumental inequality type of constraints on nonexperimental distributions.  ...  If no U variable is a descendant of any V variable, then the corresponding model is called a semi-Markovian model. In this paper, we only consider semi-Markovian models.  ... 
arXiv:1206.6829v1 fatcat:y2efxokembe5baftyqplzkcz3i

Inference in multi-agent causal models

Sam Maes, Stijn Meganck, Bernard Manderick
2007 International Journal of Approximate Reasoning  
In this article, we demonstrate the usefulness of causal Bayesian networks as probabilistic reasoning systems.  ...  The main contribution of this article is the introduction of multi-agent causal models as a way to overcome the problems in a centralized setting.  ...  Due to lack of space, we immediately introduce the theorem for semi-Markovian causal models.  ... 
doi:10.1016/j.ijar.2006.09.005 fatcat:wimkrdrjdven5fsi5wm4jrgupy

PC-Fairness: A Unified Framework for Measuring Causality-based Fairness [article]

Yongkai Wu, Lu Zhang, Xintao Wu, Hanghang Tong
2019 arXiv   pre-print
A recent trend of fair machine learning is to define fairness as causality-based notions which concern the causal connection between protected attributes and decisions.  ...  Experiments on synthetic and real-world datasets show the correctness and effectiveness of our method.  ...  On the other hand, if all exogenous variables in U are assumed to be mutually independent, then the causal model is called a Markovian model; otherwise, it is called a semi-Markovian model.  ... 
arXiv:1910.12586v1 fatcat:j3m4mzne4jb3jhdomyzwog5vky

Structural Causal Models Are (Solvable by) Credal Networks [article]

Marco Zaffalon and Alessandro Antonucci and Rafael Cabañas
2020 arXiv   pre-print
Extensive experiments show that approximate algorithms for credal networks can immediately be used to do causal inference in real-size problems.  ...  This allows to exactly map a causal model into a credal network.  ...  The SCM M is semi-Markovian if its causal diagram is acyclic.  ... 
arXiv:2008.00463v1 fatcat:kenlbdt5qjfwljgnjacwcr6iha

Diffusion Causal Models for Counterfactual Estimation [article]

Pedro Sanchez, Sotirios A. Tsaftaris
2022 arXiv   pre-print
Herein we propose Diff-SCM, a deep structural causal model that builds on recent advances of generative energy-based models.  ...  Counterfactual estimation is achieved by firstly inferring latent variables with deterministic forward diffusion, then intervening on a reverse diffusion process using the gradients of an anti-causal predictor  ...  We use an EMA rate of 0.9999 for all experiments. We use DDIM sampling for all experiments with 1000 timesteps. The same noise schedule is used for training.  ... 
arXiv:2202.10166v1 fatcat:atrw3kcgdrenhdp6dsacaio4qa

Learning and Testing Causal Models with Interventions [article]

Jayadev Acharya, Arnab Bhattacharyya, Constantinos Daskalakis, Saravanan Kandasamy
2018 arXiv   pre-print
We consider testing and learning problems on causal Bayesian networks as defined by Pearl (Pearl, 2009).  ...  We also obtain sample/time/intervention efficient algorithms for: (i) testing the identity of two unknown causal Bayesian networks on the same graph; and (ii) learning a causal Bayesian network on a given  ...  Acknowledgments We would like to thank Vasant Honavar who told us about the problems considered here and for several helpful discussions that were essential for us to complete this work.  ... 
arXiv:1805.09697v1 fatcat:ywirlx6vrjfp3jbu2viefnyfyi

The learning effects of brands: Determined through Markovian analysis of brand switching

Dilip Roy
2009 Journal of Management Research  
This paper aims at determining the learning effects of brands using Markovian analysis.  ...  For demonstration purpose, a set of information, available in the literature, on brand switching in the Indian oral care market has been used and learning effects have been empirically determined along  ...  Lawrence(1975) and Wierenga(1974) made use of learning model in different contexts. However, Leefiand and Boostra (1982) indicated a few limitations of the learning model.  ... 
doi:10.5296/jmr.v1i2.42 fatcat:q25wnnhaejesnei6oulwzqzav4

Causal Expectation-Maximisation [article]

Marco Zaffalon and Alessandro Antonucci and Rafael Cabañas
2021 arXiv   pre-print
Structural causal models are the basic modelling unit in Pearl's causal theory; in principle they allow us to solve counterfactuals, which are at the top rung of the ladder of causation.  ...  This appears to be a consequence of the fact, proven in this paper, that causal inference is NP-hard even in models characterised by polytree-shaped graphs.  ...  So the EMCC could be a significant step to make causal inference widely applicable in machine learning.  ... 
arXiv:2011.02912v3 fatcat:anmc3kc2dvf2hcaenswxlrzuia

Inferring Hidden Statuses and Actions in Video by Causal Reasoning

Amy Fire, Song-Chun Zhu
2017 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW)  
In this paper, we extend the Causal And-Or Graph (C-AOG) to a sequential model representing actions and their effects on objects over time, and we build a probability model for it.  ...  Our results demonstrate the effectiveness of reasoning with causality over time.  ...  A hidden semi-Markov model [16] can accommodate the non-Markovian duration terms while enforcing consistency.  ... 
doi:10.1109/cvprw.2017.13 dblp:conf/cvpr/FireZ17 fatcat:h3pphgkz5vhx5owc7hkhycusza

Minimum Cost Intervention Design for Causal Effect Identification [article]

Sina Akbari, Jalal Etesami, Negar Kiyavash
2022 arXiv   pre-print
Pearl's do calculus is a complete axiomatic approach to learn the identifiable causal effects from observational data.  ...  When such an effect is not identifiable, it is necessary to perform a collection of often costly interventions in the system to learn the causal effect.  ...  Terminology & Problem Description We briefly introduce the notations used in this paper 1 . We use the structural causal model framework of (Pearl et al., 2000) in this work.  ... 
arXiv:2205.02232v1 fatcat:ycfs2eyhhfg6znl5mdwe3bzkuy
« Previous Showing results 1 — 15 out of 884 results