Learning one-hidden-layer neural networks with landscape design

Keyphrases

• Neural Network 100%
• Global Minimum 100%
• Hidden Layer 100%
• Landscape Design 100%
• Local Minima 33%
• Analytic Formula 33%
• Finite Sample 33%
• Nonconvex Objective Function 33%
• Gaussian Distribution 33%
• Stochastic Gradient Descent 33%
• Squared Loss 33%
• Full Rank 33%
• Weight Matrix 33%
• Noise Vectors 33%
• Population Risk 33%
• Complexity Results 33%
• Low-rank Tensor 33%
• Ranked Positional Weight 33%
• Sample Complexity 33%

Mathematics

• Neural Network 100%
• Analytic Formula 50%
• Gaussian Distribution 50%
• Objective Function 50%
• Local Minimum 50%
• Rank Tensor 50%
• Weight Matrix 50%
• Stochastic Property 50%
• Convex 50%