arXiv subjects

Omiros Papaspiliopoulos

Publications and source records attributed to Omiros Papaspiliopoulos.

At least 19 recordsLinked to original sources

arXiv2026

Functional Estimation of the Marginal Likelihood

We propose a framework for computing, optimizing and integrating with respect to a smooth marginal likelihood in statistical models that involve high-dimensional parameters/latent variables and continuous low-dimensional hyperparameters. The method requires samples from the posterior distribution of the parameters for different values of the hyperparameters on a simulation grid and returns inference on the marginal likelihood defined everywhere on its domain, and on its functionals. We show how the method relates to many of the methods that have been used in this context, including sequential Monte Carlo, Gibbs sampling, Monte Carlo maximum likelihood, and umbrella sampling. We establish the consistency of the proposed estimators as the sampling effort increases, both when the simulation grid is kept fixed and when it becomes dense in the domain. We showcase the approach on Gaussian process regression and classification and crossed effect models.

stat.ME↗

arXiv2025

Conjugate gradient methods for high-dimensional GLMMs

Generalized linear mixed models (GLMMs) are a widely used tool in statistical analysis. The main bottleneck of many computational approaches lies in the inversion of the high dimensional precision matrices associated with the random effects. Such matrices are typically sparse; however, the sparsity pattern resembles a multi partite random graph, which does not lend itself well to default sparse linear algebra techniques. Notably, we show that, for typical GLMMs, the Cholesky factor is dense even when the original precision is sparse. We thus turn to approximate iterative techniques, in particular to the conjugate gradient (CG) method. We combine a detailed analysis of the spectrum of said precision matrices with results from random graph theory to show that CG-based methods applied to high-dimensional GLMMs typically achieve a fixed approximation error with a total cost that scales linearly with the number of parameters and observations. Numerical illustrations with both real and simulated data confirm the theoretical findings, while at the same time illustrating situations, such as nested structures, where CG-based methods struggle.

stat.CO↗

arXiv2025

Inference for multiple treatment effects using confounder importance learning

We address modelling and computational issues for multiple treatment effect inference under many potential confounders. Our main contribution is providing a trade-off between preventing the omission of relevant confounders, while not running into an over-selection of instruments that significantly inflates variance. We propose a novel empirical Bayes framework for Bayesian model averaging that learns from data the prior inclusion probabilities of key covariates. Our framework sets a data-dependent prior that asymptotically matches the true amount of confounding in the data, as measured by a novel confounding coefficient. A key challenge is computational. We develop fast algorithms, using an exact gradient of the marginal likelihood that has linear cost in the number of covariates, and a variational counterpart. Our framework uses widely-used ingredients and largely existing software, and it is implemented within the R package mombf. We illustrate our work with two applications. The first is the association between salary variation and discriminatory factors. The second, that has been debated in previous works, is the association between abortion policies and crime. Our approach provides insights that differ from previous analyses especially in situations with weaker treatment effects.

stat.ME↗

arXiv2024

Partially factorized variational inference for high-dimensional mixed models

While generalized linear mixed models are a fundamental tool in applied statistics, many specifications, such as those involving categorical factors with many levels or interaction terms, can be computationally challenging to estimate due to the need to compute or approximate high-dimensional integrals. Variational inference is a popular way to perform such computations, especially in the Bayesian context. However, naive use of such methods can provide unreliable uncertainty quantification. We show that this is indeed the case for mixed models, proving that standard mean-field variational inference dramatically underestimates posterior uncertainty in high-dimensions. We then show how appropriately relaxing the mean-field assumption leads to methods whose uncertainty quantification does not deteriorate in high-dimensions, and whose total computational cost scales linearly with the number of parameters and observations. Our theoretical and numerical results focus on mixed models with Gaussian or binomial likelihoods, and rely on connections to random graph theory to obtain sharp high-dimensional asymptotic analysis. We also provide generic results, which are of independent interest, relating the accuracy of variational inference to the convergence rate of the corresponding coordinate ascent algorithm that is used to find it. Our proposed methodology is implemented in the R package, see https://github.com/mgoplerud/vglmer . Numerical results with simulated and real data examples illustrate the favourable computation cost versus accuracy trade-off of our approach compared to various alternatives.

stat.ME↗

arXiv2023

Scalable Bayesian computation for crossed and nested hierarchical models

We develop sampling algorithms to fit Bayesian hierarchical models, the computational complexity of which scales linearly with the number of observations and the number of parameters in the model. We focus on crossed random effect and nested multilevel models, which are used ubiquitously in applied sciences. The posterior dependence in both classes is sparse: in crossed random effects models it resembles a random graph, whereas in nested multilevel models it is tree-structured. For each class we identify a framework for scalable computation, building on previous work. Methods for crossed models are based on extensions of appropriately designed collapsed Gibbs samplers, where we introduce the idea of local centering; while methods for nested models are based on sparse linear algebra and data augmentation. We provide a theoretical analysis of the proposed algorithms in some simplified settings, including a comparison with previously proposed methodologies and an average-case analysis based on random graph theory. Numerical experiments, including two challenging real data analyses on predicting electoral results and real estate prices, compare with off-the-shelf Hamiltonian Monte Carlo, displaying drastic improvement in performance.

stat.CO↗

arXiv2021

Bayesian prediction of jumps in large panels of time series data

We take a new look at the problem of disentangling the volatility and jumps processes of daily stock returns. We first provide a computational framework for the univariate stochastic volatility model with Poisson-driven jumps that offers a competitive inference alternative to the existing tools. This methodology is then extended to a large set of stocks for which we assume that their unobserved jump intensities co-evolve in time through a dynamic factor model. To evaluate the proposed modelling approach we conduct out-of-sample forecasts and we compare the posterior predictive distributions obtained from the different models. We provide evidence that joint modelling of jumps improves the predictive ability of the stochastic volatility models.

q-fin.ST↗

arXiv2020

Exact inference for a class of non-linear hidden Markov models on general state spaces

Exact inference for hidden Markov models requires the evaluation of all distributions of interest - filtering, prediction, smoothing and likelihood - with a finite computational effort. This article provides sufficient conditions for exact inference for a class of hidden Markov models on general state spaces given a set of discretely collected indirect observations linked non linearly to the signal, and a set of practical algorithms for inference. The conditions we obtain are concerned with the existence of a certain type of dual process, which is an auxiliary process embedded in the time reversal of the signal, that in turn allows to represent the distributions and functions of interest as finite mixtures of elementary densities or products thereof. We describe explicitly how to update recursively the parameters involved, yielding qualitatively similar results to those obtained with Baum--Welch filters on finite state spaces. We then provide practical algorithms for implementing the recursions, as well as approximations thereof via an informed pruning of the mixtures, and we show superior performance to particle filters both in accuracy and computational efficiency. The code for optimal filtering, smoothing and parameter inference is made available in the Julia package DualOptimalFiltering.

stat.CO↗

arXiv2019

Dimension-Robust MCMC in Bayesian Inverse Problems

The methodology developed in this article is motivated by a wide range of prediction and uncertainty quantification problems that arise in Statistics, Machine Learning and Applied Mathematics, such as non-parametric regression, multi-class classification and inversion of partial differential equations. One popular formulation of such problems is as Bayesian inverse problems, where a prior distribution is used to regularize inference on a high-dimensional latent state, typically a function or a field. It is common that such priors are non-Gaussian, for example piecewise-constant or heavy-tailed, and/or hierarchical, in the sense of involving a further set of low-dimensional parameters, which, for example, control the scale or smoothness of the latent state. In this formulation prediction and uncertainty quantification relies on efficient exploration of the posterior distribution of latent states and parameters. This article introduces a framework for efficient MCMC sampling in Bayesian inverse problems that capitalizes upon two fundamental ideas in MCMC, non-centred parameterisations of hierarchical models and dimension-robust samplers for latent Gaussian processes. Using a range of diverse applications we showcase that the proposed framework is dimension-robust, that is, the efficiency of the MCMC sampling does not deteriorate as the dimension of the latent state gets higher. We showcase the full potential of the machinery we develop in the article in semi-supervised multi-class classification, where our sampling algorithm is used within an active learning framework to guide the selection of input data to manually label in order to achieve high predictive accuracy with a minimal number of labelled data.

stat.ME↗

arXiv2019

Bayesian forecasting of electoral outcomes with new parties' competition

This paper proposed a methodology to forecast electoral outcomes using the result of the combination of a fundamental model and a model-based aggregation of polls. We propose a Bayesian hierarchical structure for the fundamental model that synthesises data at the provincial, regional and national level. We use a Bayesian strategy to combine the fundamental model with the information coming for recent polls. This model can naturally be updated every time new information, for instance a new poll, becomes available. This methodology is well suited to deal with increasingly frequent situations in which new political parties enter an electoral competition, although our approach is general enough to accommodate any other electoral situation. We illustrate the advantages of our method using the 2015 Spanish Congressional Election in which two new parties ended up receiving 30\% of the votes. We compare the predictive performance of our model versus alternative models. In general the predictions of our model outperform the alternative specifications, including hybrid models that combine fundamental and polls models. Our predictions are, in relative terms, particularly accurate in predicting the seats obtained by each political party.

stat.AP↗

arXiv2018

Scalable inference for crossed random effects models

We analyze the complexity of Gibbs samplers for inference in crossed random effect models used in modern analysis of variance. We demonstrate that for certain designs the plain vanilla Gibbs sampler is not scalable, in the sense that its complexity is worse than proportional to the number of parameters and data. We thus propose a simple modification leading to a collapsed Gibbs sampler that is provably scalable. Although our theory requires some balancedness assumptions on the data designs, we demonstrate in simulated and real datasets that the rates it predicts match remarkably the correct rates in cases where the assumptions are violated. We also show that the collapsed Gibbs sampler, extended to sample further unknown hyperparameters, outperforms significantly alternative state of the art algorithms.

stat.CO↗

arXiv2018

Auxiliary gradient-based sampling algorithms

We introduce a new family of MCMC samplers that combine auxiliary variables, Gibbs sampling and Taylor expansions of the target density. Our approach permits the marginalisation over the auxiliary variables yielding marginal samplers, or the augmentation of the auxiliary variables, yielding auxiliary samplers. The well-known Metropolis-adjusted Langevin algorithm (MALA) and preconditioned Crank-Nicolson Langevin (pCNL) algorithm are shown to be special cases. We prove that marginal samplers are superior in terms of asymptotic variance and demonstrate cases where they are slower in computing time compared to auxiliary samplers. In the context of latent Gaussian models we propose new auxiliary and marginal samplers whose implementation requires a single tuning parameter, which can be found automatically during the transient phase. Extensive experimentation shows that the increase in efficiency (measured as effective sample size per unit of computing time) relative to (optimised implementations of) pCNL, elliptical slice sampling and MALA ranges from 10-fold in binary classification problems to 25-fold in log-Gaussian Cox processes to 100-fold in Gaussian process regression, and it is on par with Riemann manifold Hamiltonian Monte Carlo in an example where the latter has the same complexity as the aforementioned algorithms. We explain this remarkable improvement in terms of the way alternative samplers try to approximate the eigenvalues of the target. We introduce a novel MCMC sampling scheme for hyperparameter learning that builds upon the auxiliary samplers. The MATLAB code for reproducing the experiments in the article is publicly available and a Supplement to this article contains additional experiments and implementation details.

stat.ML↗

arXiv2017

A note on MCMC for nested multilevel regression models via belief propagation

In the quest for scalable Bayesian computational algorithms we need to exploit the full potential of existing methodologies. In this note we point out that message passing algorithms, which are very well developed for inference in graphical models, appear to be largely unexplored for scalable inference in Bayesian multilevel regression models. We show that nested multilevel regression models with Gaussian errors lend themselves very naturally to the combined use of belief propagation and MCMC. Specifically, the posterior distribution of the regression parameters conditionally on covariance hyperparameters is a high-dimensional Gaussian that can be sampled exactly (as well as marginalized) using belief propagation at a cost that scales linearly in the number of parameters and data. We derive an algorithm that works efficiently even for conditionally singular Gaussian distributions, e.g., when there are linear constraints between the parameters at different levels. We show that allowing for such non-invertible Gaussians is critical for belief propagation to be applicable to a large class of nested multilevel models. From a different perspective, the methodology proposed can be seen as a generalization of forward-backward algorithms for sampling to multilevel regressions with tree-structure graphical models, as opposed to single-branch trees used in classical Kalman filter contexts.

stat.CO↗

arXiv2017

Scalable Bayesian variable selection and model averaging under block orthogonal design

We propose a scalable algorithmic framework for exact Bayesian variable selection and model averaging in linear models under the assumption that the Gram matrix is block-diagonal, and as a heuristic for exploring the model space for general designs. In block-diagonal designs our approach returns the most probable model of any given size without resorting to numerical integration. The algorithm also provides a novel and efficient solution to the frequentist best subset selection problem for block-diagonal designs. Posterior probabilities for any number of models are obtained by evaluating a single one-dimensional integral that can be computed upfront, and other quantities of interest such as variable inclusion probabilities and model averaged regression estimates by carrying out an adaptive, deterministic one-dimensional numerical integration. The overall computational cost scales linearly with the number of blocks, which can be processed in parallel, and exponentially with the block size, rendering it most adequate in situations where predictors are organized in many moderately-sized blocks. For general designs, we approximate the Gram matrix by a block-diagonal using spectral clustering and propose an iterative algorithm that capitalizes on the block-diagonal algorithms to explore efficiently the model space. All methods proposed in this article are implemented in the R library mombf.

stat.CO↗

arXiv2016

Conjugacy properties of time-evolving Dirichlet and gamma random measures

We extend classic characterisations of posterior distributions under Dirichlet process and gamma random measures priors to a dynamic framework. We consider the problem of learning, from indirect observations, two families of time-dependent processes of interest in Bayesian nonparametrics: the first is a dependent Dirichlet process driven by a Fleming-Viot model, and the data are random samples from the process state at discrete times; the second is a collection of dependent gamma random measures driven by a Dawson-Watanabe model, and the data are collected according to a Poisson point process with intensity given by the process state at discrete times. Both driving processes are diffusions taking values in the space of discrete measures whose support varies with time, and are stationary and reversible with respect to Dirichlet and gamma priors respectively. A common methodology is developed to obtain in closed form the time-marginal posteriors given past and present data. These are shown to belong to classes of finite mixtures of Dirichlet processes and gamma random measures for the two models respectively, yielding conjugacy of these classes to the type of data we consider. We provide explicit results on the parameters of the mixture components and on the mixing weights, which are time-varying and drive the mixtures towards the respective priors in absence of further data. Explicit algorithms are provided to recursively compute the parameters of the mixtures. Our results are based on the projective properties of the signals and on certain duality properties of their projections.

math.ST↗

arXiv2015

Exact sampling of diffusions with a discontinuity in the drift

We introduce exact methods for the simulation of sample paths of one-dimensional diffusions with a discontinuity in the drift function. Our procedures require the simulation of finite-dimensional candidate draws from probability laws related to those of Brownian motion and its local time and are based on the principle of retrospective rejection sampling. A simple illustration is provided.

stat.ME↗

arXiv2015

Towards automatic calibration of the number of state particles within the SMC$^2$ algorithm

SMC$^2$ is an efficient algorithm for sequential estimation and state inference of state-space models. It generates $N_θ$ parameter particles $θ^{m}$, and, for each $θ^{m}$, it runs a particle filter of size $N_{x}$ (i.e. at each time step, $N_{x}$ particles are generated in the state space $\mathcal{X}$). We discuss how to automatically calibrate $N_{x}$ in the course of the algorithm. Our approach relies on conditional Sequential Monte Carlo updates, monitoring the state of the pseudo random number generator and on an estimator of the variance of the unbiased estimate of the likelihood that is produced by the particle filters, which is obtained using nonparametric regression techniques. We observe that our approach is both less CPU intensive and with smaller Monte Carlo errors than the initial version of SMC$^2$.

stat.CO↗

arXiv2014

Filtering hidden Markov measures

We consider the problem of learning two families of time-evolving random measures from indirect observations. In the first model, the signal is a Fleming--Viot diffusion, which is reversible with respect to the law of a Dirichlet process, and the data is a sequence of random samples from the state at discrete times. In the second model, the signal is a Dawson--Watanabe diffusion, which is reversible with respect to the law of a gamma random measure, and the data is a sequence of Poisson point configurations whose intensity is given by the state at discrete times. A common methodology is developed to obtain the filtering distributions in a computable form, which is based on the projective properties of the signals and duality properties of their projections. The filtering distributions take the form of mixtures of Dirichlet processes and gamma random measures for each of the two families respectively, and an explicit algorithm is provided to compute the parameters of the mixtures. Hence, our results extend classic characterisations of the posterior distribution under Dirichlet process and gamma random measures priors to a dynamic framework.

math.ST↗

arXiv2014

Optimal filtering and the dual process

We link optimal filtering for hidden Markov models to the notion of duality for Markov processes. We show that when the signal is dual to a process that has two components, one deterministic and one a pure death process, and with respect to functions that define changes of measure conjugate to the emission density, the filtering distributions evolve in the family of finite mixtures of such measures and the filter can be computed at a cost that is polynomial in the number of observations. Special cases of our framework include the Kalman filter, and computable filters for the Cox-Ingersoll-Ross process and the one-dimensional Wright-Fisher process, which have been investigated before. The dual we obtain for the Cox-Ingersoll-Ross process appears to be new in the literature.

math.ST↗