> ML_ALGORITHM // HAMILTONIAN-MONTE-CARLO-NUTS_v1.0
Hamiltonian Monte Carlo & No-U-Turn Sampler (HMC / NUTS)
The state-of-the-art MCMC algorithm that uses physical Hamiltonian dynamics and automatic trajectory stopping (NUTS) to sample efficiently from high-dimensional posteriors.
Markov Chain Monte Carlo Samplingbayesian-probabilistichigh-intrinsicsmall (<1k)
Back to All AlgorithmsComputational Complexity
Training Complexity:O(chains * samples * leapfrog_steps * grad_eval)
Inference Complexity:O(samples)
Hardware Profile
CPU Friendly:Yes
Requires GPU:No
Memory Footprint:low
Interpretability & Data
Interpretability Tier:high-intrinsic
Training Data Needs:small (<1k)
Interpretability Assessment
NUTS automatically calibrates trajectory lengths to eliminate arbitrary tuning parameters.
Suitable Tasks & Supported Modalities
Suitable Tasks:
bayesian inferencehierarchical statistical modeling
Supported Modalities:
tabular
Implementing Libraries
Foundational Literature
Common Pitfalls & Warnings
- Divergent transitions when sampling from high-curvature regions (Neal's Funnel); requires non-centered parameterization re-expression