We prove a conjecture about approximating Gaussian Processes on one dimension.
problem Computational scaling issues with Gaussian Processes on one dimension.
method Developed a new family of state-space models (LEG) to approximate any stationary GP on one dimension.
result Proved that any stationary GP on one dimension can be approximated using the LEG family.
New LVMs optimize any exponential family distribution without specific assumptions.
problem Optimizing latent variable models with non-Gaussian observables.
method Generic optimization using EM approach for exponential family distributions.
result Concise parameter update equations applicable to various data types.
We describe \textit{deep exponential families} (DEFs), a class of latent variable models that are inspired by the hidden structures used in deep neural networks. DEFs capture a hierarchy of dependencies between latent variables, and are easily generalized to many settings through exponential families. We perform infere…
Develops a new method for nonlinear dimension reduction using random features.
problem Statistical challenges in generalizing Gaussian process-based latent variable models to non-Gaussian data.
method Random feature latent variable models (RFLVMs) that approximate nonlinear relationships with linear functions of random features.
result RFLVMs produce comparable results to state-of-the-art methods on various data types.
A new method infers neural trajectories in real-time, improving experimental design.
problem Real-time inference of neural trajectories for immediate feedback.
method Exponential family variational Kalman filter (eVKF) for online learning.
result eVKF achieves competitive performance on synthetic and real-world data.
EFA extends self-attention to handle mixed data types and dynamic relevance.
problem Handling high-dimensional, mixed data types with dynamic relevance.
method Probabilistic generative model using self-attention and latent factor model.
result EFA consistently outperforms existing models in complex latent structure capture and reconstruction.
New method for fitting graphical models with latent variables using regularized conditional likelihood.
problem Graphical modeling with latent variables and confounding dependencies.
method Regularized conditional likelihood for exponential family graphical models.
result Framework applicable to broader settings without knowing latent variables' distribution.
The paper introduces a new method for graph embedding using exponential family distributions.
problem Representing networks in a low dimensional latent space for various applications.
method Introduces the exponential family graph embedding model, generalizing random walk-based techniques to exponential family conditional distributions.
result The proposed techniques outperform existing methods in link prediction and node classification tasks.
This paper analyzes VAE approximation errors in conditional exponential families.
problem Posterior collapse and approximation errors in VAEs.
method Analysis of ELBO objective and conditional exponential families.
result The ELBO optimizer pulls away from the likelihood optimizer towards a consistent subset of models.
CDEFs reduce model complexity and uncover time correlations.
problem Model complexity and data efficiency in probabilistic modeling.
method Builds on deep exponential families, ties weights for reduced parameters.
result CDEFs uncover time correlations with fewer parameters.
New model identifies regimes in non-stationary data.
problem Identifying latent regimes in non-stationary systems with instantaneous effects.
method Identifiable Markov Switching Models with exponential family noise.
result Established identifiability of latent regimes and causal structures.
The Kontsevich integral of a knot is a powerful invariant which takes values in an algebra of trivalent graphs with legs. Given a Lie algebra, the Kontsevich integral determines an invariant of knots (the so-called colored Jones function) with values in the symmetric algebra of the Lie algebra. Recently A. Kricker and …
Deep equilibrium models estimate latent variables from data.
problem Estimating latent variables from data.
method Generalized exponential family models, deep equilibrium networks.
result Deep equilibrium models solve MAP estimates for latent and transformation parameters.
NatPN provides fast, accurate uncertainty estimation for exponential family distributions.
problem Uncertainty in machine learning models.
method NatPN uses Normalizing Flows to fit a single density in a latent space, updating predictions based on likelihood.
result NatPN delivers competitive performance in classification, regression, and count prediction tasks.
Generalizes Lefschetz fibrations with rational homology disk smoothings.
problem Understanding rational homology disk smoothings of surface singularities.
method Introduces a genus to generic fibers of Lefschetz fibrations.
result Families of relations in mapping class groups represent smoothings.
In this note we study the Seifert rational homology spheres with two complementary legs, i.e. with a pair of invariants whose fractions add up to one. We give a complete classification of the Seifert manifolds with 3 exceptional fibers and two complementary legs which bound rational homology balls. The result translate…
EM algorithm converges in KL divergence for exponential families via mirror descent.
problem Lack of understanding of EM's non-asymptotic convergence properties.
method Viewing EM as a mirror descent algorithm, showing convergence rates in KL divergence.
result KL divergence rates for EM in exponential families, invariant to parametrization.
This paper shows how to perform likelihood inference for complex graphical models efficiently.
problem Intractable normalizing constants in fully and partially observed exponential family graphical models.
method Using a technique from Geyer (1991), the paper estimates the normalizing constant and its gradient.
result Full likelihood-based analysis is feasible and computationally efficient for these models.
Market portfolio decomposed into body and tail legs
problem Separation of market portfolio into body and tail legs
method Dynamic value-weighted body and tail legs
result Recombination identity holds for all models
Deep generative models are commonly used for generating images and text. Interpretability of these models is one important pursuit, other than the generation quality. Variational auto-encoder (VAE) with Gaussian distribution as prior has been successfully applied in text generation, but it is hard to interpret the mean…
New method for fast inference in diffusion models.
problem Intractable probabilistic inference in diffusion models.
method Variational Gaussian Process, exponential family description, convex optimization.
result Improved fast algorithm for learning model parameters.
Study decomposes market portfolio into body and tail legs, revealing systematic differences.
problem Understanding the relationship between body and tail components in market portfolios.
method Decomposes CRSP market portfolio into body and tail legs, analyzes their recombination identity.
result Recombination identity holds for all models but not for all, indicating systematic differences.
Generative models unify heterogeneous data for multimodal fusion.
problem Learning effective representations of mixed numerical and categorical data.
method Bayesian approach with exponential family distributions and Laplace-Bernstein approximation.
result Generative models enable fusion of multimodal data for various machine learning tasks.
Legged robots pose one of the greatest challenges in robotics. Dynamic and agile maneuvers of animals cannot be imitated by existing methods that are crafted by humans. A compelling alternative is reinforcement learning, which requires minimal craftsmanship and promotes the natural evolution of a control policy. Howeve…
Fast approximate inference for non-Gaussian data.
problem Efficient inference for non-Gaussian data.
method Laplace Matching for fast approximate inference in latent Gaussian models.
result Achieves high approximation quality with low computational cost.
Generalizes moment-matching for exponential families with conditioning or hidden data.
problem Generalizing moment-matching conditions for exponential families with conditioning or hidden data.
method First-principles explanation and self-contained derivation of generalized moment-matching conditions.
result Derives generalized moment-matching conditions for conditional exponential families and hidden data.
HiPPO framework optimizes memory compression for sequential data.
problem Incremental representation of cumulative history in sequential data.
method Optimal polynomial projections for online function approximation.
result HiPPO-LegS achieves state-of-the-art accuracy on MNIST.
The study explores generalized divergences and exponential families with a focus on sufficient conditions and laws of large numbers.
problem Generalization of Kullback-Leibler divergence and exponential families.
method Investigation of (h,τ)-divergence and (h,τ)-exponential families, definition of (h,τ)-dependence, proof of law of large numbers. result Sufficient condition for (h,τ)-divergence to induce Hessian structure on (h,τ)-exponential family, proof of law of large numbers. Paper introduces kernel deformed exponential families for sparse continuous attention.
problem Creating efficient attention mechanisms for sparse data.
method Developed kernel deformed exponential families, theoretically and experimentally.
result Kernel deformed exponential families can attend to multiple compact regions of data.
Word embeddings are a powerful approach for capturing semantic similarity among terms in a vocabulary. In this paper, we develop exponential family embeddings, a class of methods that extends the idea of word embeddings to other types of high-dimensional data. As examples, we studied neural data with real-valued observ…
Paper analyzes latent space geometry in generative models using Fisher information.
problem Understanding the structure of latent spaces in generative models.
method Reconstructs Fisher information metric from generated samples and posterior distribution.
result Reveals fractal structure and abrupt changes in Fisher metric at phase boundaries.
We study topology of configuration spaces of planar linkages having one leg of variable length. Such telescopic legs are common in modern robotics where they are used for shock absorbtion and serve a variety of other purposes. Using a Morse theoretic technique, we compute explicitly, in terms of the metric data, the Be…
New framework for task-independent legged locomotion.
problem Building stable legged locomotion systems in robotics.
method Task-independent spiking central pattern generator using learning methods.
result Robotic legged locomotion at different speeds and within the same gait cycle.
Develops polynomial diffusion models for multi-factor commodity futures dynamics.
problem Modeling futures prices using latent state variables for short and long-term stochastic factors.
method Polynomial diffusion models to incorporate non-linear effects, two filtering methods for estimation.
result Accurate estimation of futures prices despite parameter identification issues in polynomial diffusion models.
Study shows Seifert fibered spaces don't bound rational homology balls.
problem Understanding when Seifert fibered spaces bound rational homology balls.
method Analyzes Seifert fibered spaces with different conditions and orientations.
result Characterizes conditions for Seifert fibered spaces to bound rational homology balls.
We study parameter inference in large-scale latent variable models. We first propose an unified treatment of online inference for latent variable models from a non-canonical exponential family, and draw explicit links between several previously proposed frequentist or Bayesian methods. We then propose a novel inference…
We generalize the stochastic block model to the important case in which edges are annotated with weights drawn from an exponential family distribution. This generalization introduces several technical difficulties for model estimation, which we solve using a Bayesian approach. We introduce a variational algorithm that …
The versatility of exponential families, along with their attendant convexity properties, make them a popular and effective statistical model. A central issue is learning these models in high-dimensions, such as when there is some sparsity pattern of the optimal parameter. This work characterizes a certain strong conve…
Tests factor models by decomposing market into body and tail legs, revealing inconsistent results.
problem Inconsistency between factor models and market behavior.
method Decomposes market into body and tail legs, testing factor models at daily and monthly frequencies.
result q5 model shows inconsistent results, with negative body and positive tail alphas at all split ratios.
The study establishes a criterion for the holomorphy of curvature in smooth webs and applies it to dual webs of homogeneous foliations.
problem Establishing conditions for the holomorphy of curvature in smooth webs and their duals.
method Developed an effective criterion for the holomorphy of curvature in smooth d-webs and applied it to dual webs of homogeneous foliations. result Characterized the holomorphy of the curvature of dual webs of homogeneous foliations on PC2. Study compares short vs long strategies for equity factors, finds short strategy better.
problem Determining the best market-neutral implementation of equity factors.
method Revisited the relative predictability of short and long legs, diversification, and costs.
result Long-Short implementation yields superior risk-adjusted returns compared to Hedged Long-Only.
We provide a classification of graphical models according to their representation as subfamilies of exponential families. Undirected graphical models with no hidden variables are linear exponential families (LEFs), directed acyclic graphical models and chain graphs with no hidden variables, including Bayesian networks …
New insights into natural exponential families improve regret bounds for bandit problems.
problem Improving regret bounds for bandit problems with subexponential tails.
method Proving self-concordance for natural exponential families and applying to bandits.
result Optimistic algorithms for generalized linear bandits have second-order regret bounds that are free of an exponential dependence on problem parameters.
Bayesian network models with latent variables are widely used in statistics and machine learning. In this paper we provide a complete algebraic characterization of Bayesian network models with latent variables when the observed variables are discrete and no assumption is made about the state-space of the latent variabl…
We consider the problem of extracting a low-dimensional, linear latent variable structure from high-dimensional random variables. Specifically, we show that under mild conditions and when this structure manifests itself as a linear space that spans the conditional means, it is possible to consistently recover the struc…
EFDA extends LDA to non-Gaussian models using exponential families.
problem Classifying non-Gaussian data with LDA's limitations.
method EFDA uses exponential families to derive closed-form estimators for natural parameters and a linear decision rule.
result EFDA matches LDA's accuracy while reducing ECE by 2-6x, proving asymptotic calibration and efficiency.
A new approach to denoising using optimal transport theory.
problem Improving latent variable recovery from noisy observations.
method Inspired by optimal transport theory, a new denoising method is developed.
result The new denoising method can recover latent variables from marginal distributions and posterior means.
A new GNN module learns geometric scattering features for better graph classification and feature exploration.
problem Learning long-range graph relations and extracting meaningful features from graphs.
method Proposes a learnable geometric scattering (LEGS) module in graph neural networks (GNNs), incorporating wavelet filters.
result LEGS-based GNNs outperform existing methods in graph classification and feature extraction tasks.