Making inferences from data streams is a pervasive problem in many modern data analysis applications. But it requires to address the problem of continuous model updating and adapt to changes or drifts in the underlying data generating distribution. In this paper, we approach these problems from a Bayesian perspective c…
On-device research index
arXiv research
A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
168,742 papers · 148 categories
Trend · papers per month
48 results for “non-conjugate priors”
Thompson sampling has impressive empirical performance for many multi-armed bandit problems. But current algorithms for Thompson sampling only work for the case of conjugate priors since these algorithms require to infer the posterior, which is often computationally intractable when the prior is not conjugate. In this …
Latent Gaussian models (LGMs) are widely used in statistics and machine learning. Bayesian inference in non-conjugate LGMs is difficult due to intractable integrals involving the Gaussian prior and non-conjugate likelihoods. Algorithms based on variational Gaussian (VG) approximations are widely employed since they str…
Proposes a non-conjugate model selection method for chain event graphs.
problem Existing model selection algorithms for chain event graphs rely on conjugate priors, which is unrealistic for many real-world applications.
method Proposes a mixture modelling approach to model selection in chain event graphs that does not rely on conjugacy.
result The proposed method is more scalable and robust than existing algorithms.
Gaussian Process Factor Analysis (GPFA) has been broadly applied to the problem of identifying smooth, low-dimensional temporal structure underlying large-scale neural recordings. However, spike trains are non-Gaussian, which motivates combining GPFA with discrete observation models for binned spike count data. The dra…
Exploratory cancer drug studies test multiple tumor cell lines against multiple candidate drugs. The goal in each paired (cell line, drug) experiment is to map out the dose-response curve of the cell line as the dose level of the drug increases. We propose Bayesian Tensor Filtering (BTF), a hierarchical Bayesian model …
We show that if a link has a closed -braid representative admitting non-degenerate exchange move, an exchange move that does not obviously preserve the conjugacy class, has infinitely many non-conjugate closed -braid representatives.
Hierarchical beta process has found interesting applications in recent years. In this paper we present a modified hierarchical beta process prior with applications to hierarchical modeling of multiple data sources. The novel use of the prior over a hierarchical factor model allows factors to be shared across different …
The stochastic variational inference (SVI) paradigm, which combines variational inference, natural gradients, and stochastic updates, was recently proposed for large-scale data analysis in conjugate Bayesian models and demonstrated to be effective in several problems. This paper studies a family of Bayesian latent vari…
Natural Gradients in Practice: Non-Conjugate Variational Inference in Gaussian Process Modelsstat.ML
The natural gradient method has been used effectively in conjugate Gaussian process models, but the non-conjugate case has been largely unexplored. We examine how natural gradients can be used in non-conjugate stochastic settings, together with hyperparameter learning. We conclude that the natural gradient can signific…
Bayesian tree ensemble model for estimating treatment effects in high-dimensional survival data.
problem Estimating heterogeneous treatment effects in censored survival data with many covariates.
method Developed a Bayesian tree ensemble model with a horseshoe prior for adaptive shrinkage.
result Accurately estimates treatment effects in high-dimensional covariate spaces and non-linear functions.
New method speeds up inference for non-conjugate Gaussian processes.
problem Inference for non-conjugate Gaussian processes is slow and unreliable.
method Automated augmented conjugate inference method that constructs auxiliary variables to make the model conditionally conjugate.
result Our method is up to two orders of magnitude faster and more robust than existing methods.
Inference of latent feature models in the Bayesian nonparametric setting is generally difficult, especially in high dimensional settings, because it usually requires proposing features from some prior distribution. In special cases, where the integration is tractable, we can sample new feature assignments according to …
In a recent preprint Yael Karshon showed that there exist non-conjugate tori in a group of symplectomorphisms of a Hirzebruch surface. She counted them in terms of the cohomology class of the symplectic structure. We show that a similar phenomenon exists in the contactomorphism groups of pre-quantum circle bundles over…
BONG optimizes Bayesian inference online with natural gradient descent.
problem Sequential Bayesian inference in online settings.
method Bayesian online natural gradient (BONG) approach based on variational Bayes.
result BONG outperforms other online VB methods in non-conjugate settings.
The paper provides guarantees for a tangent transform algorithm in logistic regression models.
problem Finding theoretical guarantees for statistical optimality and algorithmic convergence in non-conjugate models.
method Exploiting convex duality and minorizing the marginal likelihood, the paper derives non-asymptotic upper bounds and convergence guarantees for a tangent transform algorithm in logistic regression models.
result The tangent transform algorithm is shown to be locally asymptotically stable without assumptions on the data-generating process.
Paper extends multi-task Gaussian Cox processes for heterogeneous tasks.
problem Modeling multiple heterogeneous correlated tasks jointly.
method Data augmentation and mean-field approximation for non-conjugate Bayesian inference.
result Demonstrates improved performance and inference on synthetic and real data.
Efficient Bayesian variable selection for binomial and negative binomial data.
problem Computational challenges in Bayesian variable selection for complex models.
method Tempered Gibbs Sampling and MCMC scheme.
result Demonstrated effectiveness on cancer data with thousands of covariates.
We prove that under fairly general conditions an iterated exchange move gives infinitely many non-conjugate braids. As a consequence, every knot has infinitely many conjugacy classes of n-braid representations if and only if it has one admitting an exchange move.
New methods accelerate NCGP inference by trading computation for uncertainty.
problem Prohibitively expensive exact inference in NCGPs for large datasets.
method Iterative methods explicitly modeling approximation error, leveraging parallel computing.
result Significant acceleration of posterior inference compared to baselines.
Bayesian method tackles variable selection in high-dimensional data.
problem Challenges in Bayesian variable selection with large P.
method Efficient MCMC scheme with sublinear cost per iteration, extended to generalized linear models.
result Demonstrated effectiveness on cancer and maize genomic data.
Study on Metropolis-within-Gibbs schemes for high-dimensional Bayesian models.
problem Improving the scalability of MCMC methods for complex Bayesian models.
method Relating convergence properties to conditional conductance for non-conjugate hierarchical models.
result Established dimension-free convergence results for Metropolis-within-Gibbs schemes.
Unified approach to Bayesian inference with guarantees on covariance matrices.
problem Approximate Bayesian inference with PSD guarantees.
method Bayes-Newton methods extending Newton's method for optimisation.
result Novel algorithms with PSD covariance matrices.
We show that any two non-conjugate points on a forward or backward complete connected Finsler manifold can be joined by infinitely many geodesics which are not covered by finitely many closed ones, provided that the Betti numbers of the based loop space grow unbounded.
Improves hyperparameter learning in GP models with non-conjugate likelihoods.
problem Hyperparameter learning entangled with approximate inference in GP models.
method Hybrid training procedure combining VI for inference and EP-like marginal likelihood approximation for hyperparameter learning.
result Empirically demonstrates the effectiveness of the proposed training procedure across various data sets.
Graph Gaussian processes use Matérn models for better function learning.
problem Lack of Gaussian process models for graph input spaces.
method Stochastic partial differential equation characterization of Matérn Gaussian processes.
result Graph Matérn Gaussian processes inherit properties of Euclidean and Riemannian models and can be trained efficiently.
A conjugate Bayesian method detects change points in Hawkes processes efficiently.
problem Non-conjugacy between Hawkes process likelihood and prior causes inefficiency in change point detection.
method Data augmentation to propose a conjugate Bayesian two-step change point detection method.
result The conjugate method is more accurate and efficient than non-conjugate methods.
We prove the Morse relations for the set of all geodesics connecting two non-conjugate points on a class of globally hyperbolic Lorentzian manifolds. We overcome the difficulties coming from the fact that the Morse index of every geodesic is infinite, and from the lack of the Palais-Smale condition, by using the Morse …
This paper proposes a method to approximate non-Gaussian likelihoods in Gaussian Processes.
problem Approximating non-Gaussian likelihoods in Gaussian Processes.
method Proposes a piece-wise constant approximation for the inverse-link function.
result Yields a closed form solution for the SVGP lower bound.
SING improves state inference in latent SDE models for better drift function estimation.
problem Intractable posterior inference in latent SDE models.
method Natural gradient variational inference.
result SING provides faster and more reliable inference in latent SDE models.
This is an addendum to arXiv: 0810.5376. We show, using our methods and an auxiliary result of Bestvina-Bromberg-Fujiwara, that a finitely generated group with infinitely many pairwise non-conjugate homomorphisms to a mapping class group virtually acts non-trivially on an -tree, and, if it is finitely presented, it…
cvHM framework speeds up GP inference for neural spike train analysis.
problem Scalability issue in approximate inference for latent GP models.
method cvHM framework using Hida-Matérn kernels and conjugate computation variational inference (CVI).
result Linear time inference for latent neural trajectories.
The Hodge series of a finite matrix group is the generating function for invariant exterior forms of specified order and degree. Lauret, Miatello, and Rossetti gave examples of pairs of non-conjugate cyclic groups having the same Hodge series; the corresponding space forms are isospectral for the Laplacian on p-forms f…
Ensemble++ uses shared-factor ensembles to scale Thompson Sampling for linear and nonlinear bandits.
problem Computational challenges in Thompson Sampling for large-scale or non-conjugate settings.
method Ensemble++ with shared-factor architecture and random linear combinations.
result Ensemble++ achieves comparable regret to exact Thompson Sampling with significantly smaller ensemble sizes.
New subgroups of mapping class groups constructed for infinite-type surfaces.
problem Constructing new subgroups of mapping class groups for infinite-type surfaces.
method Utilization of special homeomorphisms called shift maps and multipush maps.
result Countably (and uncountably in certain cases) many non-conjugate embeddings of subgroups into mapping class groups.
Fibrations on Handlebodiesmath.GT
Uncountably many fibrations found on genus 2 handlebody.
problem Finding fibrations on specific 3-manifolds.
method Constructing fibrations with Cantor tree fibers.
result Uncountably many fibrations with non-conjugate monodromies.
We obtain a result about the existence of only a finite number of geodesics between two fixed non-conjugate points in a Finsler manifold endowed with a convex function. We apply it to Randers and Zermelo metrics. As a by-product, we also get a result about the finiteness of the number of lightlike and timelike geodesic…
We use some Lie group theory and Budney's unitarization of the Lawrence-Krammer representation, to prove that for generic parameters of definite form the image of the representation (also on certain types of subgroups) is dense in the unitary group. This implies that, except possibly for closures of full-twist braids, …
We develop methods for efficient amortized approximate Bayesian inference over posterior distributions of probabilistic clustering models, such as Dirichlet process mixture models. The approach is based on mapping distributed, symmetry-invariant representations of cluster arrangements into conditional probabilities. Th…
It has been known since the time of Nielsen that the mapping class group of a surface of genus and one puncture acts faithfully by homeomorphisms on the circle. In this note, we show that this standard representation of the mapping class group is not rigid, precisely, if is a…
A construction of representations of 3-manifold groups into PU(2,1) through Lefschetz fibrationsmath.GT
New representations of 3-manifold groups into complex hyperbolic space found.
problem Finding representations of 3-manifold groups into complex hyperbolic spaces.
method Using Lefschetz fibrations and orbifold fundamental groups of branched coverings of the projective plane.
result Infinitely many non-conjugate representations discovered.
The paper analyzes symmetries of Vaidya-Bonner geodesics.
problem Investigating invariance properties of Vaidya-Bonner geodesics.
method Classification of Lie point symmetries and Noether symmetries, determination of optimal system of subalgebras.
result Determination of optimal system of subalgebras for Vaidya-Bonner geodesics.
An Index Theory for Paths that are Solutions of a Class of Strongly Indefinite Variational Problemsmath.DG
We prove a generalized version of the Morse index theorem for geodesics endowed with a non positive definite metric tensor (semi-Riemannian manifolds). We apply the result to obtain lower estimates on the number of geodesics joining two fixed non conjugate points in certain classes of manifolds. More specifically, we c…
The paper disproves a conjecture about isomorphic subgroups in finite groups.
problem The existence of non-isomorphic subgroups that are isomorphic in extensions of finite groups.
method Constructing extensions of finite groups to show non-isomorphic pre-images of subgroups.
result Subgroups of finite groups that are isomorphic in extensions are not conjugate.
In this note, we give an explicit counterexample to the simple loop conjecture for representations of surface groups into PSL(2,R). Specifically, we show that for any surface with negative Euler characteristic and genus at least 1, there are uncountably many non-conjugate, non-injective homomorphisms of its fundamental…
Improves SVGP methods for faster and more accurate Gaussian process inference.
problem Efficient non-conjugate Gaussian process inference.
method Dual parameterization of SVGP methods using site parameters.
result Faster and more accurate inference with tighter evidence lower bound.
We describe a simple fundamental domain for the holonomy group of the boundary unipotent spherical CR uniformization of the figure eight knot complement, and deduce that small deformations of that holonomy group (such that the boundary holonomy remains parabolic) also give a uniformization of the figure eight knot comp…
The main goal of this paper is a detailed study of asymptotic cones of the mapping class groups. In particular, we prove that every asymptotic cone of a mapping class group has a bi-Lipschitz equivariant embedding into a product of real trees, sending limits of hierarchy paths onto geodesics, and with image a median su…