Research map / Learning theory and algorithms / Sample complexity
Most Neural Networks Are Almost Learnable
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 |
|---|---|---|---|
| Approximate hidden activations with low-degree Hermite polynomials | Random Xavier networks of constant depth · Any input distribution on the sphere | learning loss: at most ε | Helped |
| Use sigmoid or ReLU-like activations with exponentially decreasing polynomial approximation error | Polynomial-learning PTAS · Random Xavier networks of constant depth | time and sample complexity: (\bar d)^{O\left(\left(\log\left(\frac{14(L+1)^3}{\epsilon}\right)\right)^{(i-1)}\right)} | Helped |
Closest papers by meaning
- Computational Complexity of Learning Neural Networks: Smoothness and Degeneracy (NeurIPS 2023 · similarity 0.88)
- SGD Learns the Conjugate Kernel Class of the Network (NeurIPS 2017 · similarity 0.86)
- An Exact Poly-Time Membership-Queries Algorithm for Extracting a Three-Layer ReLU Network (ICLR 2023 · similarity 0.86)
- Lower Bounds on the Depth of Integral ReLU Neural Networks via Lattice Polytopes (ICLR 2023 · similarity 0.86)
- Random Sparse Lifts: Construction, Analysis and Convergence of finite sparse networks (ICLR 2024 · similarity 0.85)
- On the Hardness of Training Deep Neural Networks Discretely (AAAI 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.