Research map / Learning theory and algorithms / Kernels and regression theory
Optimal Approximation and Generalization Errors for Deep Convolutional Neural Networks
Jinxin Wang, Shao-Bo Lin
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add max-pooling to zero-padded deep CNN (ablation) | eDCNN · f1 with noise tau=0.5 | test RMSE: 0.1482 -> 0.1100 | Improved |
| add max-pooling to zero-padded deep CNN (ablation) | eDCNN · f1 with noise tau=0.01 | test RMSE: 0.0302 -> 0.0129 | Improved |
| add max-pooling to zero-padded deep CNN (ablation) | eDCNN · f2 filter length sweep, L=8 | test RMSE | Improved |
| add max-pooling to reduce run-to-run variance (ablation) | eDCNN · f1 approximation (10 trials) | variance of test RMSE | Improved |
| add a fully connected layer after convolutions (ablation) | cDCNN · f2 approximation | test RMSE | Improved |
| widen DFCN or add fully connected layer under high noise (ablation) | DFCN, cDCNN-fc · noise tau=0.3 or 0.5 | test RMSE | Hurt |
Closest papers by meaning
- Understanding Generalization and Optimization Performance of Deep CNNs (ICML 2018 · similarity 0.88)
- Optimal Rates for Generalization of Gradient Descent for Deep ReLU Classification (NeurIPS 2025 · similarity 0.87)
- Approximation and non-parametric estimation of ResNet-type convolutional neural networks (ICML 2019 · similarity 0.87)
- Generalization Error Bounds of Gradient Descent for Learning Over-Parameterized Deep ReLU Networks (AAAI 2020 · similarity 0.86)
- Minimax optimality of convolutional neural networks for infinite dimensional input-output problems and separation from kernel methods (ICLR 2024 · similarity 0.86)
- Understanding deep learning requires rethinking generalization (ICLR 2017 · similarity 0.86)
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.
