Stan models are compiled to generative languages and extended with new features.
problem Lack of direct support for variational inference and deep models in Stan.
method Comprehensive compilation scheme to convert Stan models to generative languages, and extension of Stan with new features.
result NumPyro backend yields a 2.3x speedup compared to Stan in geometric mean over 26 benchmarks.
SlicStan improves Stan's usability and efficiency.
problem Stan's block syntax sacrifices usability for scalability.
method Formalized Stan, introduced SlicStan with compositional syntax and flexible functions.
result SlicStan facilitates better code reuse and abstraction.
Variational inference is a scalable technique for approximate Bayesian inference. Deriving variational inference algorithms requires tedious model-specific calculations; this makes it difficult to automate. We propose an automatic variational inference algorithm, automatic differentiation variational inference (ADVI). …
DynamicPPL speeds up probabilistic modeling in Julia.
problem Developing and executing complex dynamic probabilistic models efficiently.
method Modular Julia library with a DSL, tracing data structure, and contextual dispatch.
result Achieves computational performance close to or better than Stan.
AutoStan improves Bayesian models via predictive feedback.
problem Improving Bayesian models written in Stan.
method Iterative improvement of Stan models using NLPD and sampler diagnostics feedback.
result AutoStan can autonomously improve diverse Bayesian models across various structures.
Fenrir efficiently estimates Bayesian MLN-DLMs for scalable inference.
problem Computational challenges in Bayesian MLN-DLMs for longitudinal count compositional data.
method Novel algorithm for MAP estimation and accurate posterior marginal approximation.
result Fenrir can be three orders of magnitude more efficient than Stan.
Stan fits econophysics models to market data, providing a rigorous statistical approach.
problem Quantifying the role of complex, heterogeneous actors in market dynamics.
method Bayesian modeling using Stan, focusing on likelihood-based fitting.
result Stan's likelihood-based approach offers advantages over moment matching methods.
ADVI simplifies model fitting for probabilistic models.
problem Fitting complex models to large datasets is computationally challenging.
method Automatic differentiation variational inference (ADVI).
result ADVI automatically derives efficient variational inference algorithms.
New model learns relative importance of multiple seasonal patterns in time series data.
problem Complex seasonal patterns in business time series data.
method Mixed hierarchical seasonality (MHS) model using Stan.
result Significant improvements in prediction error and predictive density compared to existing models.
We solve Bayesian PCA's rotational symmetry issue by rotation-invariant parameterization.
problem Bayesian PCA's rotational symmetry complicates inference and interpretation.
method Rotation-invariant Householder parameterization using random matrix theory.
result Efficient rotation-invariant probabilistic PCA implementation.
BREAD evaluates MCMC inference quality using bidirectional sampling.
problem Measuring the reliability of MCMC inference is hard, especially in black box methods.
method BREAD combines bidirectional sampling and AIS to bound KL divergence.
result BREAD provides a protocol to validate MCMC-based inference relevance.
A simple approach speeds up probabilistic programming models.
problem Efficiently implementing probabilistic programming in deep learning.
method Embedding probabilistic programming into TensorFlow with a single random variable abstraction.
result Optimal linear speedup from 1 to 256 TPUv2 chips, and 100x speedup on GPUs.
TrustVI is a fast second-order algorithm for black-box variational inference.
problem Efficiently optimizing variational distributions in complex models.
method Trust-region optimization with minibatch reparameterization.
result TrustVI converges faster and finds better variational distributions than alternatives.
Edward is a flexible probabilistic programming language that integrates with TensorFlow.
problem Probabilistic programming challenges in flexibility and efficiency.
method Compositional representations of random variables and inference, flexible inference methods, integration with TensorFlow.
result Edward achieves significant speedups in probabilistic modeling and inference.
Research forecasts electricity spot prices using stochastic volatility models.
problem Forecasting day-ahead electricity prices in a spot market.
method Exploring and enriching a baseline stochastic volatility model with exogenous regressors.
result A better fitting model confirmed by out-of-sample forecasts.
New neural network models speed up Bayesian multilevel modeling.
problem Complex computational challenges in Bayesian multilevel modeling.
method Probabilistic neural network architectures that leverage multilevel model factorization.
result Efficient posterior inference on unseen datasets with near-instant results.
New method for efficient probabilistic inference using masked language modeling.
problem Efficient posterior inference in probabilistic programs with many hyper-parameters.
method Formulate inference as masked language modeling, train a neural network to unmask random values.
result Foundation posterior for zero-shot inference and fine-tuning across a range of programs.
Paper advances black-box VI using flows and Monte-Carlo methods.
problem Improving automatic posterior inference in black-box VI.
method Combines normalizing flows, Monte-Carlo methods, and optimization considerations.
result Significant improvement in state-of-the-art variational inference.
Bayesian framework improves minority class performance in class-imbalanced data.
problem Class imbalance in predictive toxicology models.
method Weighted likelihood approach modifying likelihood function weights inversely proportional to class proportions.
result Improves balanced accuracy and sensitivity for minority class (toxic compounds).
A novel VAE called πVAE models complex data with stochastic processes.
problem Efficient inference of stochastic processes for Bayesian deep learning with big data.
method πVAE is a finitely exchangeable, Kolmogorov consistent, continuous stochastic process VAE.
result πVAE achieves state-of-the-art performance in tasks like spatial interpolation.
NeuralProphet improves forecast accuracy by 55-92% for short-term forecasts.
problem Challenges in explainable, scalable forecasting for business decisions.
method Hybrid framework combining classical methods and deep learning, with auto-regression and covariate modules.
result NeuralProphet outperforms Prophet on real-world datasets and short-term forecasts.
Automates Bayesian signal processing algorithm design using factor graphs.
problem Designing efficient Bayesian signal processing algorithms.
method Factor graph approach and ForneyLab tool for automated inference.
result ForneyLab outperforms competitors in automated inference for state-space models.
The empirical practice of using factor models to adjust for shared, unobserved confounders, Z, in observational settings with multiple treatments, A, is widespread in fields including genetics, networks, medicine, and politics. Wang and Blei (2019, WB) formalizes these procedures and develops the …
Bayesian Neural Networks combine neural networks and stochastic models for probabilistic predictions.
problem Creating probabilistic guarantees for neural network predictions.
method Combines neural networks and stochastic models, focusing on posterior distribution generation.
result BNNs provide probabilistic guarantees and distribution of learned parameters.
Type system captures CI relationships for probabilistic models.
problem Challenges in inference for models with mixed discrete and continuous parameters.
method Information flow type system for probabilistic programming.
result Well-typed programs guarantee certain CI relationships.
Proposes a new bankruptcy prediction model using Bayesian framework with expert knowledge.
problem Bankruptcy prediction using accounting ratios and statistical models.
method Generative modeling with Expert Bayesian framework.
result Model is either comparable or superior to existing methods with lower False Positive Rate.
Let Λ be the limit set of a conformal dynamical system, i.e. a Kleinian group acting on either finite- or infinite-dimensional real Hilbert space, a conformal iterated function system, or a rational function. We give an easily expressible sufficient condition, requiring that the limit set is not too much bigger than …