arXiv research

Paper reconciles RCM and SCM frameworks for causal inference.

It is common practice in using regression type models for inferring causal effects, that inferring the correct causal relationship requires extra covariates are included or ``adjusted for''. Without performing this adjustment erroneous causal effects can be inferred. Given this phenomenon it is common practice to inclu…

Generic groups can't move spaces but have rich actions.

Estimates counterfactual outcomes linking observed and unobserved data.

ProPublica's analysis of recidivism predictions produced by Correctional Offender Management Profiling for Alternative Sanctions (COMPAS) software tool for the task, has shown that the predictions were racially biased against African American defendants. We analyze the COMPAS data using a causal reformulation of the un…

Starting from the recent classification of quotients of Freund--Rubin backgrounds in string theory of the type AdS_{p+1} x S^q by one-parameter subgroups of isometries, we investigate the physical interpretation of the associated quotients by discrete cyclic subgroups. We establish which quotients have well-behaved cau…

New framework infers causal shifts in event sequences under out-of-domain interventions.

As virtually all aspects of our lives are increasingly impacted by algorithmic decision making systems, it is incumbent upon us as a society to ensure such systems do not become instruments of unfair discrimination on the basis of gender, race, ethnicity, religion, etc. We consider the problem of determining whether th…

This paper uses machine learning to estimate how different types of crashes affect highway traffic.

Paper presents a method to align unpaired samples across different modalities.

New method improves causal inference by estimating complex treatment effects with active learning.

This paper improves credit line impact analysis by considering spending as a distribution.

We interpret black box predictive models using causal attribution.

We investigate the problem of estimating the causal effect of a treatment on individual subjects from observational data, this is a central problem in various application domains, including healthcare, social sciences, and online advertising. Within the Neyman Rubin potential outcomes model, we use the Kullback Leibler…

This paper addresses clustering with missing data using Rubin's rules.

Rubin LSST DESC uses AI/ML for dark energy research.

Model criticism is usually carried out by assessing if replicated data generated under the fitted model looks similar to the observed data, see e.g. Gelman, Carlin, Stern, and Rubin [2004, p. 165]. This paper presents a method for latent variable models by pulling back the data into the space of latent variables, and c…

New method refines prediction intervals for individual treatment effects using cross-world correlation.

Bayes-UCBVI tackles reinforcement learning with a new upper confidence bound method.

Extends boundary estimates for Monge-Ampère equations in polygonal domains.

Framework improves PV forecasting by accounting for missing data uncertainty.

We classify Freund-Rubin backgrounds of eleven-dimensional supergravity of the form AdS_4 x X^7 which are at least half BPS; equivalently, smooth quotients of the round 7-sphere by finite subgroups of SO(8) which admit an (N>3)-dimensional subspace of Killing spinors. The classification is given in terms of pairs consi…

We give general classification and structure theorems for actions of groups of homeomorphisms and diffeomorphisms on manifolds, reminiscent of classical results for actions of (locally) compact groups. This gives a negative answer to Ghys' "extension problem" for diffeomorphisms of manifolds with boundary, as well as a…

The volumes, spectra and geodesics of a recently constructed infinite family of five-dimensional inhomogeneous Einstein metrics on the two $S^3$ bundles over $S^2$ are examined. The metrics are in general of cohomogeneity one but they contain the infinite family of homogeneous metrics $T^{p,1}$. The geodesic flow is sh…

We study the eleven dimensional supergravity equations which describe a low energy approximation to string theories and are related to M-theory under the AdS/CFT correspondence. These equations take the form of a non-linear differential system, on $\mathbb{B}^7\times\mathbb{S}^4$ with the characteristic degeneracy at t…

New model improves data augmentation for causal tasks.

We quantify causal bias in continuous treatment settings.

New geometric flow equations describe how space-time dimensions change.

Deep Causal Graphs model complex causal relationships using neural networks.

New causal distances improve evaluation of causal discovery algorithms.

Bayesian model averaging improves causal effect estimation by averaging over multiple models.

DoWhy-GCM extends causal inference in graphical models for diverse queries.

InGRA models for efficient Granger causality learning in multivariate time series.

ABCI infers causal models and queries simultaneously using Bayesian active learning.

Graph neural networks help infer causal effects from partially observable data.

TRAM-DAG models bridge interpretability and flexibility in causal modeling.

Flow models recover causal transformations from observational data and a valid ordering.

Identification of causal direction between a causal-effect pair from observed data has recently attracted much attention. Various methods based on functional causal models have been proposed to solve this problem, by assuming the causal process satisfies some (structural) constraints and showing that the reverse direct…

Proposes Causal Loss to improve machine learning models' causal inference.

Causal KL improves on existing metrics for evaluating causal models.

We consider joint estimation of multiple graphical models arising from heterogeneous and high-dimensional observations. Unlike most previous approaches which assume that the cluster structure is given in advance, an appealing feature of our method is to learn cluster structure while estimating heterogeneous graphical m…

CRN learns causal models using neural networks, scaling with variables and leveraging prior knowledge.

Generates realistic time-series data from causal models.

Probabilistic models can handle causal inference without special tools.

Causal normalizing flows recover causal models from observational data.

Paper axiomatizes interventional probability distributions.

This work restricts hidden cardinality in causal models to infer causal relations.

Proposes TNCM-VAE for generating causal financial time series.