Learning Single-Index Models with Shallow Neural Networks

Keyphrases

• Shallow Neural Network 100%
• Single-index Model 100%
• High Dimension 33%
• Ability to Learn 33%
• Learning Algorithm 33%
• Specific Structure 33%
• Near-optimal 33%
• Gradient Flow 33%
• Class Function 33%
• Natural Class 33%
• Neural Studies 33%
• Sample Complexity 33%
• Link Function 33%
• Statistical Aspects 33%
• Target Function 33%
• Low-dimensional Structures 33%
• Optimization Landscape 33%
• Semiparametric Methods 33%
• One Dimensional Projection 33%
• Generalization Guarantees 33%
• Random Initialization 33%
• Shallow Network 33%
• Optimal Sample Complexity 33%

Computer Science

• Single Index Model 100%
• Semiparametric Method 33%

Mathematics

• Model Index 100%