Home

Group

Research

Software

Teaching

Blog

Activities

October 2024 - Keynote Talk at Challenges in Categorical Data Analysis workshop, LSE

September 2024 - APTS Week 4 on Causal Inference, Oxford

July 2024 - General Chair of UAI

June 2024 - Speaker at Frontiers in Statistical Science Workshop, UCL

February 2024 - Statistics Seminar, King's College London

January 2024 - Seminar of the Statistics Section, University of Copenhagen

November 2023 - Algebraic Economics Workshop, IMSI, Chicago

October 3rd, 2023 - RSS Discussion Meeting and DeMO Presentation - RSS, London

more

 

Preprints

a Outcome-Informed Weighting for Robust ATE Estimation
(with Linying Yang)
b Simulating Longitudinal Data from Marginal Structural Models
(with Xi Lin, Daniel de Vassimon Manela, Chase Mathis, and Jens Magelund Tarp)
c Testing Generalizability in Causal Inference
(with Daniel de Vassimon Manela and Linying Yang)
d Data fusion for efficiency gain in ATE estimation: A practical review with simulations
(with Xi Lin and Jens Magelund Tarp)
e A fast score-based search algorithm for maximal ancestral graphs using entropy
(with Zhongyi Hu)
f Regression Identifiability and Edge Interventions in Linear Structural Equation Models
(with Bohao Yao)

 

Publications

2025 Combining Experimental and Observational Data through a Power Likelihood
(with Xi Lin and Jens Magelund Tarp), Biometrics (accepted)
2024 a Parameterizing and Simulating from Causal Models (with discussion)
(with Vanessa Didelez), Journal of the Royal Statistical Society, Series B 86 (3), pp 535-568
b Towards standard imsets for maximal ancestral graphs
(with Zhongyi Hu), Bernoulli 30 (3), pp 2026-2051
c Marginal Causal Flows for Validation and Inference
(with Daniel de Vassimon Manela and Laura Battaglia), NeurIPS
d A Kernel Test for Causal Association via Noise Contrastive Backdoor Adjustment
(with Robert Hu and Dino Sejdinovic) Journal of Machine Learning Research 25 (160), pp 1-56
e Doubly Robust Kernel Statistics for Testing Distributional Treatment Effects Even Under One Sided Overlap (with Jake Fawkes, Robert Hu and Dino Sejdinovic)
Transactions of Machine Learning Research
f Toward a Complete Criterion for Value of Information in Insoluble Decision Problems
(with Ryan Carey and Sanghack Lee), Transactions of Machine Learning Research
g Identifying potential catalysts to accelerate the achievement of SDGs among adolescents living in Nigeria (with Rita Tamambang and others) Psychology, Health & Medicine 29 (4) pp 868-887
h Two are Better Than One but Three is Best (with Rita Tamambang and others)
Child Indicators Research (in press)
2023 a Nested Markov Properties for Acyclic Directed Mixed Graphs
(with Thomas Richardson, James Robins and Ilya Shpitser) Annals of Statistics 51 (1) pp 334-361
b Latent-free equivalent mDAGs
Algebraic Statistics 14 (1) pp 3–16
c PWSHAP: A Path-Wise Explanation Model for Targeted Variables
(with Lucile Ter-Minassian, Oscar Clivio, Karla Diaz-Ordaz and Chris Holmes) ICML, PMLR 202: pp 34054-34089
d Results on Counterfactual Invariance
(with Jake Fawkes) SCIS (ICML Workshop)
2022 a Selection, Ignorability and Challenges With Causal Fairness
(with Jake Fawkes and Dino Sejdinovic) CLeaR 2022
b Algebraic Properties of Gaussian HTC-identifiable Graphs
(with Bohao Yao) Algebraic Statistics 13 (1) pp 19-39
c Accelerating Progress Towards the Sustainable Development Goals for adolescents in Ghana: a cross-sectional study
(with Kwabena Kusi-Mensah and others) Psychology, Health & Medicine 27 pp 49-66
2021 a Dependency in DAG models with hidden variables
UAI-21, PMLR 161 pp 813-822
b Exploring the relationship between pain and self-harm thoughts and behaviours in young people using network analysis
(with Verena Hinze, Tamsin Ford, Bergljot Gjelsvik and Catherine Crane) Psychological Medicine, 1-10
2020 a Model selection and local geometry
Annals of Statistics, 48 (6), pp 3514-3544
b Faster Algorithms for Markov equivalence
(with Zhongyi Hu) UAI-20, PMLR 124 pp 739-748
c Dissociation in relation to other mental health conditions: An exploration using network analysis
(with Emma Černis, Anke Ehlers and Daniel Freeman) Journal of Psychiatric Research
d Comment on: Graphical models for extremes by Engelke and Hitz
Journal of the Royal Statistical Society, Series B, 82 (4), pp 919-920
2019 a Smooth, identifiable supermodels of discrete DAG models with latent variables
(with Thomas Richardson) Bernoulli, 25 (2) pp 848-876
b Maximum likelihood estimation of the Latent Class Model through model boundary decomposition
(with Elizabeth Allman and others) Journal of Algebraic Statistics, 10 (1) pp 51-84
c Adolescent Paranoia: Prevalence, Structure, and Causal Mechanisms
(with Jessica Bird and others), Schizophrenia Bulletin, 45 (5), pp 1134-1142
d Markov Properties for Mixed Graphical Models
Chapter 2 of Handbook of Graphical Models (Maathuis et al., Eds)
2018 a Margins of discrete Bayesian networks
Annals of Statistics, 46 (6A) pp 2623-2656
b Acyclic Linear SEMs Obey the Nested Markov Property
(with Ilya Shpitser and Thomas Richardson) UAI-18, (supplementary material)
c Causal Inference from Case-Control Studies
(with Vanessa Didelez) Chapter 6 of Handbook of Statistical Methods for Case-Control Studies (Borgan et al., Eds)
2017 Distributional equivalence and structure Learning for Bow-free Acyclic Path Diagrams
(with Christopher Nowzohour, Marloes H. Maathuis and Peter Bühlmann)
Electronic Journal of Statistics, 11 (2), pp 5342-5374
2016 a Graphs for margins of Bayesian networks
Scandinavian Journal of Statistics, 43 (3), pp 625-648
b Causal Inference through a Witness Protection Program
(with Ricardo Silva) Journal of Machine Learning Research 17 (56) pp 1-53
(expansion of NIPS paper below)
c One-Component Regular Variation and Graphical Modeling of Extremes
(with Adrien Hitz) Journal of Applied Probability, 53 (3), pp 733-746
2015 a Smoothness of marginal log-linear parameterizations
Electronic Journal of Statistics, 9 (1), pp 475-491
b Recovering from Selection Bias using Marginal Structure in Discrete Models
(with Vanessa Didelez), UAI-15, Advances in Causal Inference Workshop.
2014 a Markovian acyclic directed mixed graphs for discrete data
(with Thomas Richardson), Annals of Statistics, 42 (4), pp 1452-1482
b Causal Inference through a Witness Protection Program
(with Ricardo Silva) NIPS 27
c Introduction to nested Markov models
(with Ilya Shpitser, Thomas Richardson and James Robins) Behaviormetrika 41 (1) pp 3-39
d Graphical latent structure testing
Studies in Theoretical and Applied Statistics, Springer
2013 a Marginal log-linear parameters for graphical Markov models
(with Thomas Richardson), J. Roy. Statist. Soc. B, 75 (4) pp 743-768
(software for simulations and data analysis available here)
b Two algorithms for fitting constrained marginal models
(with Antonio Forcina), Computational Statistics and Data Analysis, 66 pp 1-7.
c Sparse nested Markov models with log-linear parameters
(with Ilya Shpitser, Thomas Richardson and James Robins) UAI-13, pp 576-585
d Comment on: On the application of discrete marginal graphical models, by Németh and Rudas
Sociological Methodology, 43 (1) pp 105-107
2012 a Graphical methods for inequality constraints in marginalized DAGs
22nd Workshop on Machine Learning and Signal Processing
b Parameter and Structure Learning in Nested Markov Models
(with Ilya Shpitser, Thomas Richardson and James Robins)
UAI-12, Causal Structure Learning Workshop.
2011 Transparent parametrizations of models for potential outcomes (with discussion)
(with Thomas Richardson and James Robins), Bayesian Statistics 9, pp 569-610
2010 Maximum likelihood fitting of acyclic directed mixed graphs to binary data
(with Thomas Richardson), UAI-10, pp 177-184

 

Older Preprints

2018 Towards Characterising Bayesian Network Models under Selection
(with Angelos Armen)
2016 Modeling Website Visits
(with Adrien Hitz)

 

Thesis

Parametrizations of Discrete Graphical Models, University of Washington, 2011.
Supervisor: Thomas Richardson. (this version includes some minor corrections from the original)

 

Talks

Keynote Talk at Challenges in Categorical Data Analysis workshop, LSE, October 2024

Parameterizing and Simulating from Causal Models, Discussion Meeting (video), Royal Statistical Society, 2023

Towards standard imsets for maximal ancestral graphs, Munich Miniworkshop, September 2023

Combining Randomized and Observational Datasets, JICI Workshop, UC Berkeley, September 2022

Parameterizing and Simulating from Causal Models, Biostatistics Seminar, Copenhagen University, May 2022

mDAGs Equivalent to Latent-Free DAGs, Quantum Workshop, Simons Institute, April 2022

The Inflation Technique, Algebraic Aspects of Causality RG, Simons Program on Causality, April 2022

Parameterizing and Simulating from Causal Models, Pacific Causal Inference Conference, September 2021

Faster models for Markov equvialence, Talk at ISI, July 2021

Parameterizing Causal Models, Karolinska and MRC Cambridge Seminars, June 2020

Angles and Model Selection, Technische Universität München, October 2019

Model Selection and Local Geometry - Workshop on Causal inference for complex graphical structures, Montreal, June 2018

Causal Models with Latent Variables - Quantum Networks Workshop, Oxford, August 2017

Geometry of Graphical Model Selection - ICMS, April 2017

more

 

Other Work

(not to be reproduced without appropriate citation / permission)

2009 Discussion of Generalized Additive Models with Implicit Variable Selection by Likelihood-Based Boosting, G. Tutz and H. Binder
STAT 572 Project; a version of Tutz and Binder's paper can be found here, final version published in Biometrics 62 (2006)
2009 Maximum Likelihood Estimates for Binary Random Variables on Trees via Phylogenetic Ideals
A project for STAT 538 - a version of Zwiernik and Smith's paper can be found here
2007 Rates of Convergence of Non-Parametric Maximum Likelihood Estimators via Entropy Methods