A Conceptual Introduction to Hamiltonian Monte Carlo
arXiv:1701.02434
Abstract
Hamiltonian Monte Carlo has proven a remarkable empirical success, but only recently have we begun to develop a rigorous understanding of why it performs so well on difficult problems and how it is best applied in practice. Unfortunately, that understanding is confined within the mathematics of differential geometry which has limited its dissemination, especially to the applied communities for which it is particularly important. In this review I provide a comprehensive conceptual account of these theoretical foundations, focusing on developing a principled intuition behind the method and its optimal implementations rather of any exhaustive rigor. Whether a practitioner or a statistician, the dedicated reader will acquire a solid grasp of how Hamiltonian Monte Carlo works, when it succeeds, and, perhaps most importantly, when it fails.
60 pages, 42 figures
References in corpus (1)
Cited by in corpus (158)
- dynesty: A Dynamic Nested Sampling Package for Estimating Bayesian Posteriors and Evidences
- Visualization in Bayesian workflow
- B-PINNs: Bayesian Physics-Informed Neural Networks for Forward and Inverse PDE Problems with Noisy Data
- Sparsity information and regularization in the horseshoe and other shrinkage priors
- Validating Bayesian Inference Algorithms with Simulation-Based Calibration
- Bayesian Item Response Modeling in R with brms and Stan
- Global Analysis of the Gravitational Wave Signal from Galactic Binaries
- The Degree of Alignment Between Circumbinary Disks and Their Binary Hosts
- The Pantheon+ Analysis: Evaluating Peculiar Velocity Corrections in Cosmological Analyses with Nearby Type Ia Supernovae
- Nested Sampling Methods
- Principal component analysis and sparse polynomial chaos expansions for global sensitivity analysis and model calibration: application to urban drainage simulation
- HypoSVI: Hypocenter inversion with Stein variational inference and Physics Informed Neural Networks
- A Hierarchical Bayesian SED Model for Type Ia Supernovae in the Optical to Near-Infrared
- Quantifying dimensionality: Bayesian cosmological model complexities
- zeus: A Python implementation of Ensemble Slice Sampling for efficient Bayesian parameter inference
- Bayesian inference of network structure from unreliable data
- NeuTra-lizing Bad Geometry in Hamiltonian Monte Carlo Using Neural Transport
- Machine Learning Gravitational Waves from Binary Black Hole Mergers
- Quantum Hamiltonian-Based Models and the Variational Quantum Thermalizer Algorithm
- Preconditioned BFGS-based Uncertainty Quantification in elastic Full Waveform Inversion
- Rates and delay times of type Ia supernovae in the Dark Energy Survey
- Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicality
- CHEOPS Precision Phase Curve of the Super-Earth 55 Cnc e
- Bayesian inference for gravitational waves from binary neutron star mergers in third-generation observatories
- Warm Jupiters in TESS Full-Frame Images: A Catalog and Observed Eccentricity Distribution for Year 1
- Modified Hamiltonian Monte Carlo for Bayesian inference
- Implicitly Adaptive Importance Sampling
- Reconciling Kinetic and Equilibrium Models of Bacterial Transcription
- ZhuSuan: A Library for Bayesian Deep Learning
- A fast multi-fidelity method with uncertainty quantification for complex data correlations: Application to vortex-induced vibrations of marine risers
- Calibration of Deep Probabilistic Models with Decoupled Bayesian Neural Networks
- Auto-Differentiable Spectrum Model for High-Dispersion Characterization of Exoplanets and Brown Dwarfs
- Leave-One-Out Cross-Validation for Bayesian Model Comparison in Large Data
- Global Convergence of Stochastic Gradient Hamiltonian Monte Carlo for Non-Convex Stochastic Optimization: Non-Asymptotic Performance Bounds and Momentum-Based Acceleration
- Efficient sampling of constrained high-dimensional theoretical spaces with machine learning
- Likelihood-free MCMC with Amortized Approximate Ratio Estimators
- Kernel-based Prediction of Non-Markovian Time Series
- Practical Hilbert space approximate Bayesian Gaussian processes for probabilistic programming
- A Survey of Bayesian Statistical Approaches for Big Data
- Sampling as optimization in the space of measures: The Langevin dynamics as a composite optimization problem
- Symplectic integration of learned Hamiltonian systems
- Bayesian graph convolutional neural networks via tempered MCMC
- Efficient Learning of the Parameters of Non-Linear Models using Differentiable Resampling in Particle Filters
- Volesti: Volume Approximation and Sampling for Convex Polytopes in R
- Density estimation with Gaussian processes for gravitational-wave posteriors
- Approximating Bayes in the 21st Century
- A piecewise deterministic Monte Carlo method for diffusion bridges
- Evaluating uncertainties in electrochemical impedance spectra of solid oxide fuel cells
- Pathfinder: Parallel quasi-Newton variational inference
- Does Hamiltonian Monte Carlo mix faster than a random walk on multimodal densities?
- Bayesian Variational Autoencoders for Unsupervised Out-of-Distribution Detection
- Hamiltonian Adaptive Importance Sampling
- Fast Threshold Tests for Detecting Discrimination
- Impact of using the ultra-high-energy cosmic ray arrival energies to constrain source associations
- K-CLASH: Strangulation and Ram Pressure Stripping in Galaxy Cluster Members at 0.3 < z < 0.6
- A Geometric Theory of Higher-Order Automatic Differentiation
- The Information Content of Projected Galaxy Fields
- Neural Bridge Sampling for Evaluating Safety-Critical Autonomous Systems
- Operational, gauge-free quantum tomography
- Hierarchical Bayesian CMB Component Separation with the No-U-Turn Sampler
- Bayesian Non-parametric Quantile Process Regression and Estimation of Marginal Quantile Effects
- Hamiltonian Monte Carlo with Energy Conserving Subsampling
- lgpr: An interpretable nonparametric method for inferring covariate effects from longitudinal data
- A general perspective on the Metropolis-Hastings kernel
- Computing Bayes: From Then 'Til Now'
- Emerging Directions in Geophysical Inversion
- Hamiltonian Monte Carlo using an adjoint-differentiated Laplace approximation: Bayesian inference for latent Gaussian models and beyond
- Radio Galaxy Shape Measurement with Hamiltonian Monte Carlo in the Visibility Domain
- Roll-back Hamiltonian Monte Carlo
- Truncated Log-concave Sampling with Reflective Hamiltonian Monte Carlo
- A scalable optimal-transport based local particle filter
- A deep surrogate approach to efficient Bayesian inversion in PDE and integral equation models
- Differentiable Likelihoods for Fast Inversion of 'Likelihood-Free' Dynamical Systems
- Bias correction in daily maximum and minimum temperature measurements through Gaussian process modeling
- Deep Markov Chain Monte Carlo
- Inferring Galactic magnetic field model parameters using IMAGINE - An Interstellar MAGnetic field INference Engine
- Bayesian Inference for Randomized Benchmarking Protocols
- Using a Bayesian-Inference Approach to Calibrating Models for Simulation in Robotics
- Quantum-Inspired Hamiltonian Monte Carlo for Bayesian Sampling
- Sampling and Optimization on Convex Sets in Riemannian Manifolds of Non-Negative Curvature
- Manifold lifting: scaling MCMC to the vanishing noise regime
- Hamiltonian Monte Carlo on Symmetric and Homogeneous Spaces via Symplectic Reduction
- Schr{ö}dinger-F{ö}llmer Sampler: Sampling without Ergodicity
- Bayesian Paired-Comparison with the bpcs Package
- varstan: An R package for Bayesian analysis of structured time series models with Stan
- A Condition Number for Hamiltonian Monte Carlo
- : A robust MCMC convergence diagnostic with uncertainty using decision tree classifiers
- Delayed rejection Hamiltonian Monte Carlo for sampling multiscale distributions
- Automatic Relevance Determination Bayesian Neural Networks for Credit Card Default Modelling
- A Theoretical Connection Between Statistical Physics and Reinforcement Learning
- rstap: An R Package for Spatial Temporal Aggregated Predictor Models
- Mixed Hamiltonian Monte Carlo for Mixed Discrete and Continuous Variables
- Multi-Rank Sparse and Functional PCA: Manifold Optimization and Iterative Deflation Techniques
- Selecting the Metric in Hamiltonian Monte Carlo
- Adaptive Path Sampling in Metastable Posterior Distributions
- A Probabilistic State Space Model for Joint Inference from Differential Equations and Data
- Markov Chain Monte Carlo Methods, a survey with some frequent misunderstandings
- Modeling racial/ethnic differences in COVID-19 incidence with covariates subject to non-random missingness
- Spatial Temporal Aggregated Predictors to Examine Built Environment Health Effects
- Implicit Hamiltonian Monte Carlo for Sampling Multiscale Distributions
- Understanding MCMC Dynamics as Flows on the Wasserstein Space
- Modeling Item Response Theory with Stochastic Variational Inference
- Reality-check for Econophysics: Likelihood-based fitting of physics-inspired market models to empirical data
- Subsampling MCMC - An introduction for the survey statistician
- Measure Transport with Kernel Stein Discrepancy
- Practical Bayesian System Identification using Hamiltonian Monte Carlo
- Markov Chain Monte Carlo Methods for Estimating Systemic Risk Allocations
- High-Dimensional Gaussian Process Inference with Derivatives
- An Uncertainty Quantification Method for Inexact Simulation Models
- MCMC generation of cosmological fields far beyond Gaussianity
- A Unifying and Canonical Description of Measure-Preserving Diffusions
- IMAGINE: Testing a Bayesian pipeline for Galactic Magnetic Field model optimization
- On Strongly Coupled Matrix Theory and Stochastic Quantization: A New Approach to Holographic Dualities
- Flow field tomography with uncertainty quantification using a Bayesian physics-informed neural network
- Hamiltonian Monte Carlo for Regression with High-Dimensional Categorical Data
- A splitting Hamiltonian Monte Carlo method for efficient sampling
- Correlated Dynamics in Marketing Sensitivities
- TOI-3362b: A Proto-Hot Jupiter Undergoing High-Eccentricity Tidal Migration
- A Mathematical Walkthrough and Discussion of the Free Energy Principle
- Divide-and-Conquer MCMC for Multivariate Binary Data
- Efficient Sampling from Feasible Sets of SDPs and Volume Approximation
- Probabilistic Mapping of Dark Matter by Neural Score Matching
- Variational Inference for Shrinkage Priors: The R package vir
- Couplings for Multinomial Hamiltonian Monte Carlo
- Sequential pCN-MCMC, an efficient MCMC method for Bayesian inversion of high-dimensional multi-Gaussian priors
- Bayesian Computing in the Undergraduate Statistics Curriculum
- Functional probabilistic programming for scalable Bayesian modelling
- GPU-based Parallel Computation Support for Stan
- Monte Carlo Simulations of BFSS and IKKT Matrix Models
- f-SAEM: A fast Stochastic Approximation of the EM algorithm for nonlinear mixed effects models
- Posterior-based proposals for speeding up Markov chain Monte Carlo
- Automated Augmented Conjugate Inference for Non-conjugate Gaussian Process Models
- Discrete Sampling using Semigradient-based Product Mixtures
- HMCF - Hamiltonian Monte Carlo Sampling for Fields - A Python framework for HMC sampling with NIFTy
- Maximum likelihood estimation for a stochastic SEIR system for COVID-19
- Lipschitz standardization for multivariate learning
- Efficient Semi-Implicit Variational Inference
- A Mixed-Membership Model for Social Network Clustering
- Skills to not fall behind in school
- Invexifying Regularization of Non-Linear Least-Squares Problems
- Quantized Variational Inference
- On Numerical Considerations for Riemannian Manifold Hamiltonian Monte Carlo
- Regional Topics in British Grocery Retail Transactions
- The Ball Pit Algorithm: A Markov Chain Monte Carlo Method Based on Path Integrals
- Spatio-temporal extreme event modeling of terror insurgencies
- Entropy-based adaptive Hamiltonian Monte Carlo
- Robust Particle Density Tempering for State Space Models
- A Compartment Model of Human Mobility and Early Covid-19 Dynamics in NYC
- Meta-analysis of dichotomous and ordinal tests without a gold standard
- Knowledge is reward: Learning optimal exploration by predictive reward cashing
- Teaching Uncertainty Quantification in Machine Learning through Use Cases
- Spatial modelling of COVID-19 incident cases using Richards' curve: an application to the Italian regions
- Automatically Differentiable Random Coefficient Logistic Demand Estimation
- Learning the Prediction Distribution for Semi-Supervised Learning with Normalising Flows
- Maximum conditional entropy Hamiltonian Monte Carlo sampler
- Differentiable probabilistic programming for strong gravitational lensing
- Bayesian model selection in the -open setting -- Approximate posterior inference and probability-proportional-to-size subsampling for efficient large-scale leave-one-out cross-validation
- An introduction to computational complexity in Markov Chain Monte Carlo methods