Paper proposes VAE-BPTF for better tensor factorization of sparse, imbalanced count data.
problem Inference of Bayesian Poisson-Gamma models for sparse and imbalanced count data is challenging.
method Variational auto-encoder framework with multi-layer perceptron networks for complex update information sharing and reweighting.
result VAE-BPTF outperforms current models in reconstruction errors and latent factor coherence across real-world datasets.
Method evaluates disentanglement in DLVMs, including those not aligned with latent axes.
problem Evaluate disentanglement in DLVMs, especially those not aligned with latent axes.
method Proposes a statistical method to discover generative factors of a dataset.
result Empirically demonstrates the advantage of the method on two datasets.
LPF provides formal guarantees for aggregating multi-evidence in probabilistic tasks.
problem Lack of formal guarantees for multi-evidence reasoning in AI.
method LPF uses variational autoencoders and Sum-Product Networks to aggregate evidence items.
result Proves multiple formal guarantees including calibration preservation and error decay.
New method estimates covariance in multi-view data with better accuracy and uncertainty.
problem Estimating covariance in multi-view data with shared and view-specific latent factors.
method Spectral decompositions and conditional conjugate priors for factor loadings and residual variances.
result Proves favorable asymptotic properties and excellent performance in simulations and real data.
Variational language models seek to estimate the posterior of latent variables with an approximated variational posterior. The model often assumes the variational posterior to be factorized even when the true posterior is not. The learned variational posterior under this assumption does not capture the dependency relat…
Paper addresses the disparity between sampled and mean representations in disentangled learning.
problem Disparity between sampled and mean representations in disentangled learning.
method Proposes a method to eliminate the disparity by proving and utilizing the relationship between total correlation of sampled and mean representations for multivariate normal distributions.
result Demonstrates that a factorized mean representation can have lower total correlation than the sampled representation.
We build upon probabilistic models for Boolean Matrix and Boolean Tensor factorisation that have recently been shown to solve these problems with unprecedented accuracy and to enable posterior inference to scale to Billions of observation. Here, we lift the restriction of a pre-specified number of latent dimensions by …
PPT optimizes transformer behavior by steering its latent posterior using prior samples.
problem Eliciting desired behavior from transformers without backpropagation.
method Posterior Prefix Tuning (PPT) uses predictive Monte Carlo (PMC) samples and importance sampling to optimize the latent posterior.
result PPT optimizes transformer behavior without backpropagation, achieving high utility across different utility functions.
Deep latent variable models (LVM) such as variational auto-encoder (VAE) have recently played an important role in text generation. One key factor is the exploitation of smooth latent structures to guide the generation. However, the representation power of VAEs is limited due to two reasons: (1) the Gaussian assumption…
To improve the ability of VAE to disentangle in the latent space, existing works mostly focus on enforcing independence among the learned latent factors. However, the ability of these models to disentangle often decreases as the complexity of the generative factors increases. In this paper, we investigate the little-ex…
Model for dynamic relational data with regime changes.
problem Handling abrupt changes in dynamic relational data.
method Factorized fusion shrinkage model with global-local shrinkage priors.
result Posterior distribution attains minimax optimal rate up to logarithmic factors.
Alpha signals for statistical arbitrage strategies are often driven by latent factors. This paper analyses how to optimally trade with latent factors that cause prices to jump and diffuse. Moreover, we account for the effect of the trader's actions on quoted prices and the prices they receive from trading. Under fairly…
This project compares MCMC and VI for Bayesian PMF on MovieLens.
problem Intractable posterior distribution in PMF.
method Employed MCMC and VI for Bayesian inference on MovieLens.
result VI converges faster, MCMC provides more accurate estimates.
Models for recommender systems use latent factors to explain the preferences and behaviors of users with respect to a set of items (e.g., movies, books, academic papers). Typically, the latent factors are assumed to be static and, given these factors, the observed preferences and behaviors of users are assumed to be ge…
We address a portfolio selection problem that combines active (outperformance) and passive (tracking) objectives using techniques from convex analysis. We assume a general semimartingale market model where the assets' growth rate processes are driven by a latent factor. Using techniques from convex analysis we obtain a…
A scalable factorized Gaussian process VAE for faster inference.
problem Inference bottlenecks in Gaussian process VAEs.
method Factorizes latent kernel across auxiliary features, leveraging independence.
result Significant speed-up in inference time (in theory and practice).
A gamma process dynamic Poisson factor analysis model is proposed to factorize a dynamic count matrix, whose columns are sequentially observed count vectors. The model builds a novel Markov chain that sends the latent gamma random variables at time (t−1) as the shape parameters of those at time t, which are linked …
Stochastic variational inference (SVI) is emerging as the most promising candidate for scaling inference in Bayesian probabilistic models to large datasets. However, the performance of these methods has been assessed primarily in the context of Bayesian topic models, particularly latent Dirichlet allocation (LDA). Deri…
Improved Monte-Carlo models by constraining mutual information between latent and observable variables.
problem Training density models leads to latent variables being useless.
method Weave tighter Monte-Carlo bounds with mutual information constraints.
result Improved training of models with continuous and discrete latent variables.
The Importance Weighted Auto Encoder (IWAE) objective has been shown to improve the training of generative models over the standard Variational Auto Encoder (VAE) objective. Here, we derive importance weighted extensions to AVB and AAE. These latent variable models use implicitly defined inference networks whose approx…
A new method uses hyperspherical latent spaces to disentangle data with periodic structures.
problem Disentangling data with periodic or cyclic underlying factors in Euclidean space.
method Diffusion Variational Autoencoder with a modified Evidence Lower Bound.
result The method can recover periodic true factors effectively.
EigenBayes: A fast, adaptive Bayesian shrinkage approach for high-dimensional matrix factorization
problem Choosing the latent dimension k in factor models method Adaptive spectral shrinkage and empirical Bayes calibration
result Adapts to signal-to-noise ratio and shrinks superfluous components
Disentangled representations, where the higher level data generative factors are reflected in disjoint latent dimensions, offer several benefits such as ease of deriving invariant representations, transferability to other tasks, interpretability, etc. We consider the problem of unsupervised learning of disentangled rep…
Latent variable models can be used to probabilistically "fill-in" missing data entries. The variational autoencoder architecture (Kingma and Welling, 2014; Rezende et al., 2014) includes a "recognition" or "encoder" network that infers the latent variables given the data variables. However, it is not clear how to handl…
Paper improves variational inference for complex models.
problem Improving statistical accuracy of variational inference in high-dimensional models.
method Developed a general framework for MFVI and proposed a partially grouped VI algorithm.
result Proposed algorithm works and outperforms vanilla MFVI in mixed membership stochastic blockmodel.
SPIDER uses deep neural networks for streaming tensor factorization.
problem Lack of effective approach for deep tensor factorization of streaming data.
method Bayesian neural networks with spike-and-slab prior, Taylor expansions, moment matching, and EPI framework.
result Effective incremental updates for latent factors and NN weights.
Temporal coarse-graining of multi-sector default count data generates effective correlation matrices and rank copulas.
problem Explaining the difference in default dependence between monthly and annual aggregation.
method Dynamic low-rank state-space model with AR(1) latent credit-state factors.
result Effective correlation matrices and rank copulas are generated from monthly default count data.
Bayesian hypergraph inference models disease pathways from EHR data.
problem Modeling rare diseases influenced by shared risk factors.
method Bayesian hypergraph inference framework reframing multi-disease modeling.
result Interpretable disease pathways and well-calibrated uncertainty quantification.
RVRAE combines deep learning and dynamic factor models for better stock returns prediction.
problem Improving stock returns prediction in volatile markets.
method Combines dynamic factor modeling with variational recurrent autoencoder (VRAE). Uses prior-posterior learning for optimal factor model.
result RVRAE outperforms traditional methods in predicting stock returns and estimating variances.
Optimizes variational inference for dynamic network models.
problem Estimating pairwise inner products and intercepts in dynamic latent space models.
method Structured mean-field variational inference with block coordinate ascent algorithm.
result Variational risk attains minimax optimal rate with logarithmic factor under certain conditions.
A scalable method for efficient inference in Gaussian process regression networks.
problem Intractable inference in Gaussian process regression networks (GPRN).
method Tensorization of output space, tensor/matrix-normal variational posteriors, joint optimization, and exploiting Kronecker product structure.
result Captures posterior dependencies and improves inference quality for large number of outputs.
New research shows posterior collapse in VAEs isn't just about KL-divergence.
problem Posterior collapse in Variational Autoencoders (VAEs).
method Analyzes the loss surface of deep autoencoder networks and proves the existence of bad local minima.
result Posterior collapse in VAEs is caused by bad local minima, not just KL-divergence.
KL annealing helps VAEs avoid posterior collapse and overfitting.
problem Posterior collapse and overfitting in VAEs.
method Theoretical analysis of learning dynamics with KL annealing.
result Posterior collapse is inevitable when β exceeds a threshold. This paper clarifies VAE's property through geometric and information-theoretic interpretations.
problem The transparency of VAE model is an underlying issue.
method Quantitative understanding of VAE through differential geometry and information theory.
result VAE can be mapped to an implicit isometric embedding with a scale factor derived from the posterior parameter.
BSFP method reveals latent patterns in multi-omic data for predicting lung function in HIV-associated OLD.
problem Limited understanding of multi-omic molecular phenomena and clinical outcomes in obstructive lung disease.
method Bayesian Simultaneous Factorization and Prediction (BSFP) method for multi-omic data, accommodating imputation and full posterior inference.
result BSFP reveals distinct clusters of patients with OLD and multi-omic patterns related to lung function decline.
Scalable approach for high-dimensional dynamical systems with noise filtering and parameter estimation.
problem Noise filtering and parameter estimation for high-dimensional dynamical systems.
method Flexible latent factor model with orthogonal factor loading matrix and closed-form parameter estimation.
result Substantial acceleration and higher accuracy compared to alternatives.
A new method for analyzing latent space models without reference configurations.
problem Posterior summaries of latent coordinates in Euclidean latent space models are not canonical due to likelihood invariance.
method Quotient-based posterior analysis using the centered Gram map.
result Intrinsic posterior summaries of mean structure and uncertainty can be computed directly from posterior samples.
Variational autoencoders often collapse, showing latent variables are non-identifiable.
problem Posterior collapse in variational autoencoders due to non-identifiable latent variables.
method Proves latent variable non-identifiability causes posterior collapse. Proposes latent-identifiable models using Brenier maps and input convex neural networks.
result Latent-identifiable models resolve posterior collapse and provide meaningful representations.
Bayesian non-linear latent variable modeling for complex data.
problem Inference for GPLVMs is computationally limited and often leads to overfitting or underestimates uncertainty.
method Approximate Gaussian process mappings with random Fourier features for MCMC inference.
result Generalized RFLVMs perform well on various data types and applications.
While the Matrix Generalized Inverse Gaussian (MGIG) distribution arises naturally in some settings as a distribution over symmetric positive semi-definite matrices, certain key properties of the distribution and effective ways of sampling from the distribution have not been carefully studied. In this paper…
A new method for efficient Gaussian process inference using sparse approximations.
problem Scalable and accurate inference for latent Gaussian processes.
method Variational approximation with sparse inverse Cholesky factors and double Kullback-Leibler minimization.
result The proposed method can achieve highly accurate approximations with polylogarithmic time complexity.
Amortized inference allows latent-variable models trained via variational learning to scale to large datasets. The quality of approximate inference is determined by two factors: a) the capacity of the variational distribution to match the true posterior and b) the ability of the recognition network to produce good vari…
Improved VAEs by training a contrastive prior to match posterior.
problem Prior hole problem in VAEs, leading to poor image generation.
method Introduced a contrastive energy-based prior and trained it using noise contrastive estimation.
result Significant improvement in VAE generative performance on various datasets.
An efficient way to learn deep density models that have many layers of latent variables is to learn one layer at a time using a model that has only one layer of latent variables. After learning each layer, samples from the posterior distributions for that layer are used as training data for learning the next layer. Thi…
New method improves generative model performance by fully conditioning variational posteriors.
problem Inaccurate inference due to partial conditioning of variational posteriors in sequential LVMs.
method Introduces fully-conditioned approximate posteriors to improve generative model performance.
result Improves generative modelling and multi-step prediction performance.
New model improves DNA methylation data analysis.
problem Analyzing DNA methylation data with complex distributions.
method Doubly non-central beta (DNCB) distribution for non-negative matrix factorization.
result Improves predictive performance and yields meaningful latent representations.
DDVI uses diffusion models for variational inference, improving latent variable model performance.
problem Improving variational inference in latent variable models.
method Introduces diffusion-based variational posteriors trained with a regularized ELBO.
result Outperforms alternative variational posteriors on various benchmarks and a biology task.
This work tackles posterior collapse in conditional and hierarchical VAEs.
problem Posterior collapse in VAEs leads to poor latent variable representations.
method Theoretical analysis of linear conditional and hierarchical VAEs, empirical validation.
result Theoretical and empirical evidence of posterior collapse causes in conditional and hierarchical VAEs.