Research map / Optimization / Optimizers
Learning Weight Uncertainty With Stochastic Gradient MCMC for Shape Classification
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use pSGLD instead of RMSprop | FNN · MNIST | test error (%): 1.79 -> 1.45 | Helped |
| Add Dropout to pSGLD (ablation) | FNN 400-400 · MNIST | test error (%): 1.45 -> 1.36 | No clear effect |
| Use pSGLD instead of RMSprop | FNN 800-800 · Animal | accuracy (%): 84.30± 0.55 -> 84.80± 0.35 | Helped |
| Use pSGLD instead of RMSprop | FNN · Body Shape | accuracy (%): 56.31± 2.34 -> 57.33 ± 3.06 | Helped |
| Use pSGLD instead of RMSprop | FNN · Textured Shape Geometry and Texture | accuracy (%): 37.01 -> 38.46; 49.39 -> 51.49 | Helped |
| Use pSGLD instead of RMSprop | FNN 800-800 · 20 Newsgroups | accuracy (%): 72.52 -> 73.32 | Helped |
| Use pSGLD instead of RMSprop | CNN · MNIST | test error (%): 0.65 -> 0.45; 0.62 -> 0.49 | Helped |
| Add Dropout (ablation) | pSGLD CNN · MNIST | test error (%): 0.45 -> 0.40; 0.49 -> 0.37 | Helped |
| Add Dropout (ablation) | RMSprop CNN · Caltech 101 Silhouettes | test error (%): 27.52 -> 27.43 | No clear effect |
| Add Batch Normalization (ablation) | 5-CNN · Caltech 101 Silhouettes | test error (%): 27.52 -> 24.40 | Helped |
| Use pSGLD instead of RMSprop | Vol-CNN · ModelNet40 | test error (%): 17.06 -> 16.21 | Helped |
| Add Dropout (ablation) | pSGLD View-CNN · ModelNet40 | test error (%): 15.12 -> 14.62 | Hurt |
Closest papers by meaning
- Parameter Expanded Stochastic Gradient Markov Chain Monte Carlo (ICLR 2025 · similarity 0.88)
- Learning to Explore for Stochastic Gradient MCMC (ICML 2024 · similarity 0.86)
- Cyclical Stochastic Gradient MCMC for Bayesian Deep Learning (ICLR 2020 · similarity 0.86)
- Structured Stochastic Gradient MCMC (ICML 2022 · similarity 0.85)
- MCMC Shape Sampling for Image Segmentation With Nonparametric Shape Priors (CVPR 2016 · similarity 0.85)
- High-Order Stochastic Gradient Thermostats for Bayesian Learning of Deep Models (AAAI 2016 · similarity 0.85)
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.