Research map / Optimization / Implicit bias and generalization
Globally Optimal Training of Generalized Polynomial Neural Networks with Nonlinear Spectral Methods
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 NLSM2 instead of the compared models | Cancer classification · Cancer | test accuracy: NLSM2 96.4; ReLU1 95.7; ReLU2 93.6; SVM 95.7 | Helped |
| Use NLSM2 instead of the compared models | Haberman classification · Haberman | test accuracy: NLSM2 75.4; ReLU1 70.5; ReLU2 72.1; SVM 72.1 | Helped |
| Use NLSM2 instead of the compared models | Pima classification · Pima | test accuracy: NLSM2 80.5; NLSM1 79.2; ReLU1 76.6; ReLU2 79.2; SVM 79.9 | Helped |
| Use NLSM1 or NLSM2 instead of the RBF-kernel SVM | UCI classification tasks overall · UCI datasets | test accuracy | Hurt |
| Use NLSM1 or NLSM2 instead of ReLU networks | UCI classification tasks overall · UCI datasets | test accuracy | Mixed |
| Train with the nonlinear spectral method rather than Batch-SGD | One-hidden-layer network optimization · Training experiment | training speed | Helped |
Closest papers by meaning
- Global Optimality in Neural Network Training (CVPR 2017 · similarity 0.90)
- A Convergence Analysis of Gradient Descent for Deep Linear Neural Networks (ICLR 2019 · similarity 0.89)
- Fast Convergence of Natural Gradient Descent for Over-Parameterized Neural Networks (NeurIPS 2019 · similarity 0.89)
- Globally Optimal Gradient Descent for a ConvNet with Gaussian Inputs (ICML 2017 · similarity 0.89)
- Global Optimality Conditions for Deep Neural Networks (ICLR 2018 · similarity 0.89)
- Global Minimizers of $\ell^p$-Regularized Objectives Yield the Sparsest ReLU Neural Networks (NeurIPS 2025 · similarity 0.87)
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.