Finite-sum composition optimization via variance reduced gradient descent

Keyphrases

• Gradient Descent 100%
• Finite Sum 100%
• Composition Optimization 100%
• Reduced Gradient Method 100%
• Variance-reduced 100%
• Machine Learning 50%
• Compositional Gradient 50%
• Linear Convergence 50%
• Strongly Convex Optimization 50%
• Stochastic Composite Optimization 50%
• Optimization Formula 50%
• Stochastic Variance Reduced Gradient 50%
• Learning Statistics 50%

Mathematics

• Stochastics 100%
• Variance 100%
• Finite Sum 100%
• Minimizes 33%
• Convergence Rate 33%
• Linear Convergence 33%

Computer Science

• Gradient Descent 100%
• Convex Optimization 50%
• Convergence Rate 50%
• Machine Learning 50%
• Learning System 50%