AI Paper Atlas

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

SpHMC: Spectral Hamiltonian Monte Carlo

Haoyi Xiong, Kafeng Wang, Jiang Bian, Zhanxing Zhu, Cheng-Zhong Xu, Zhishan Guo, Jun Huan

AAAI 2019 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Stochastic gradient MCMC · 5 citations (OpenAlex)

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
Disable Metropolis-Hastings correction (ablation)SpHMC · EMNIST, l2-SVMtest error (%): 15.38 -> 22.71Hurt
Disable Metropolis-Hastings correction (ablation)SpHMC · MNIST, SVMtest error (%): 3.59 -> 3.61No clear effect
Disable Metropolis-Hastings correction (ablation)SpHMC · MNIST, l2-Perceptrontest error (%): 12.54 -> 18.84Hurt
Augment with SpHMC samples instead of original data onlySVM · MNISTtest error (%): 5.96 -> 3.59Improved
Augment with SpHMC samples instead of original data onlyl2-SVM · Fashion MNISTtest error (%): 19.88 -> 15.19Improved
Augment with SpHMC samples instead of original data onlySVM · EMNISTtest error (%): 21.21 -> 13.14Improved
Use SpHMC instead of Kernel HMC for augmentationl1-Log Reg · MNISTtest error (%): 34.14 -> 5.49Improved
Use SpHMC instead of Kernel HMC for augmentationSVM · EMNISTtest error (%): 25.18 -> 13.14Improved
Use SpHMC instead of conditional GAN for augmentationl2-Log Reg · EMNISTtest error (%): 33.21 -> 14.41Improved
Use SpHMC instead of GAN for augmentationSVM · MNISTtest error (%): 5.99 -> 3.59Improved
Use SpHMC samples instead of plain compression (LASSO spectral representation)SVM · MNISTtest error (%): 3.61 -> 3.59Improved
Use SpHMC instead of compressionl2-Perceptron · Fashion MNISTtest error (%): 25.09 -> 19.82Improved

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