Research map / Graph learning / Graph self-supervised learning
Constructing Deep Neural Networks by Bayesian Network Structure Learning
Read the paperCheck 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 |
|---|---|---|---|
| Replace the MNIST-Man head with a learned B2N structure | MNIST-Man · MNIST | accuracy difference: +0.10±0.04 | Helped |
| Replace the Maxout NiN head with a learned B2N structure | Maxout NiN · SVHN | accuracy difference: -0.40±0.05 | Hurt |
| Replace the SVHN-Man head with a learned B2N structure | SVHN-Man · SVHN | accuracy difference: -0.86±0.05 | Hurt |
| Replace the VGG-16 head with a learned B2N structure | VGG-16 · CIFAR-10 | accuracy difference: +0.29±0.14 | Helped |
| Replace the WRN-40-4 head with a learned B2N structure | WRN-40-4 · CIFAR-10 | accuracy difference: +0.33±0.14 | Helped |
| Replace the VGG-16 head with a learned B2N structure | VGG-16 · CIFAR-100 | accuracy difference: +0.05±0.17 | No clear effect |
| Replace the AlexNet head with a learned B2N structure | AlexNet · ImageNet | accuracy difference: +0.00±0.03 | No clear effect |
| Replace the AlexNet head with a learned B2N structure | AlexNet · ImageNet | parameter reduction ratio: 13.3× | Helped |
| Replace the VGG-16 head with a learned B2N structure | VGG-16 · CIFAR-100 | parameter reduction ratio: 7.7× | Helped |
| Learn the network structure directly from pixels instead of using convolutional feature extraction | B2N · MNIST | accuracy: 98.75% -> 99.07% | Helped |
Closest papers by meaning
- Bayesian Graph Convolutional Neural Networks for Semi-Supervised Classification (AAAI 2019 · similarity 0.87)
- Structure-Aware Convolutional Neural Networks (NeurIPS 2018 · similarity 0.87)
- Structured Convolutions for Efficient Neural Network Design (NeurIPS 2020 · similarity 0.86)
- Parallel Bayesian Network Structure Learning (ICML 2018 · similarity 0.86)
- Differentiable Learning of Generalized Structured Matrices for Efficient Deep Neural Networks (ICLR 2024 · similarity 0.85)
- A Symmetry-Aware Exploration of Bayesian Neural Network Posteriors (ICLR 2024 · 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.