AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Parsimonious Bayesian deep networks

Mingyuan Zhou

NeurIPS 2018 · Architectures and efficiency › RNNs and normalization · approach: Theory of depth and width · 6 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
Use PBDN-AICε with Gibbs inference rather than RBF SVMbinary classification benchmarks · imageclassification error: 2.84 -> 2.30Improved
Use PBDN-AICε with Gibbs inference rather than RBF SVMbinary classification benchmarks · bananaclassification error: 10.85 -> 10.97No clear effect
Use PBDN-AICε with Gibbs inference rather than RBF SVMbinary classification benchmarks · ijcnn1classification error: 4.01 -> 4.06Hurt
Use AICε-Gibbs rather than AIC-Gibbs for depth selection (ablation)PBDN · imageclassification error: 2.36 -> 2.30Improved
Use AICε-Gibbs rather than AIC-Gibbs for depth selection (ablation)PBDN · ijcnn1classification error: 4.33 -> 4.06Improved
Use AICε-Gibbs rather than AIC-Gibbs for depth selection (ablation)PBDN · waveformclassification error: 11.42 -> 11.45No clear effect
Use four hidden layers rather than one (ablation)PBDN · bananaclassification error: 22.54 -> 10.85Improved
Use two hidden layers rather than one (ablation)PBDN · imageclassification error: 3.32 -> 2.43Improved
Increase PBDN depth from one to two layers (ablation)PBDN · breast cancerclassification error: 29.22 -> 30.26Hurt
Train a jointly optimized DNN using the architecture inferred by PBDN instead of greedy layer-wise PBDN (ablation)PBDN-inferred architectures · eight benchmark datasetsmean SVM-normalized error: 1.047, 1.011, 1.076, and 1.144Mixed
Stack CPM layers rather than use PBDN's iSHM layers (ablation)CPM · waveformclassification error: 0.131, 0.177, 0.223Hurt
Stack CPM layers rather than use PBDN's iSHM layers (ablation)CPM · imageclassification error: 0.046, 0.080, 0.216Hurt

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