AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Adaptive Bayesian Sampling with Monte Carlo EM

Anirban Roychowdhury, Srinivasan Parthasarathy

NeurIPS 2017 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Hamiltonian Monte Carlo

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Learn the mass matrix using MCEMHMC · 1D standard normalRMSE (μ): 0.0196 -> 0.0115Improved
Learn the mass matrix using MCEMSGNHT · 1D standard normalRMSE (τ): 0.0335 -> 0.0289Improved
Learn the mass matrix using MCEMSGHMC · 1D standard normalRMSE (τ): 0.1646 -> 0.2243Hurt
Learn the mass matrix using MCEMHMC · Synthetic Bayesian logistic regressionRMSE (W₀): 0.0456 -> 0.0145Improved
Learn the mass matrix using MCEMSG-NPHMC · Synthetic Bayesian logistic regressionRMSE (W₀): 0.4945 -> 0.0990Improved
Learn the mass matrix using MCEMSG-NPHMC · Synthetic Bayesian logistic regressionRMSE (W₁): 0.4263 -> 0.4229No clear effect
Learn the mass matrix using MCEMSG-NPHMC · 20-Newsgroupstest perplexity: 714 -> 712Improved
Learn the mass matrix using MCEMSGNHT · 20-Newsgroupstest perplexity: 757 -> 719Improved
Learn the mass matrix using MCEMSGHMC · Reuterstest perplexity: 996 -> 972Improved
Use the MCEM mass-learning framework instead of the Riemannian samplerSG-NPHMC · 20-Newsgroupstime per iteration: 0.410s -> 0.049sImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy