Research map / Learning theory and algorithms / Sample complexity
Tight Sample Complexity of Learning One-hidden-layer Convolutional Neural Networks
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 |
|---|---|---|---|
| Use approximate gradient updates instead of Double Convotron | One-hidden-layer CNN · Synthetic ReLU networks, Gaussian inputs | parameter estimation error | Mixed |
| Use approximate gradient updates instead of Double Convotron | One-hidden-layer CNN · Synthetic sigmoid and hyperbolic tangent networks, Gaussian inputs | convergence | Helped |
| Use approximate gradient updates with random initialization | One-hidden-layer CNN · One-hidden-layer CNN, standard Gaussian inputs, non-overlapping filters | sample complexity: \tilde O ( (k+r)\cdot \epsilon^{-2} ) vs. \tilde O(k^2r\cdot \epsilon^{-2}) | Helped |
| Apply approximate gradient descent to all six tested activation and network-size settings | Approximate GD · Synthetic ReLU, sigmoid, and hyperbolic tangent networks | convergence | Helped |
Closest papers by meaning
- How Many Samples are Needed to Estimate a Convolutional Neural Network? (NeurIPS 2018 · similarity 0.91)
- On the Complexity of Learning Neural Networks (NeurIPS 2017 · similarity 0.88)
- Gradient Descent Learns One-hidden-layer CNN: Don’t be Afraid of Spurious Local Minima (ICML 2018 · similarity 0.88)
- The Sample Complexity of One-Hidden-Layer Neural Networks (NeurIPS 2022 · similarity 0.88)
- When is a Convolutional Filter Easy to Learn? (ICLR 2018 · similarity 0.87)
- On the Local Complexity of Linear Regions in Deep ReLU Networks (ICML 2025 · 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.