Biased stochastic first-order methods for conditional stochastic optimization and applications in meta learning

Keyphrases

• Stochastic Gradient Descent 100%
• Stochastic Optimization 100%
• Stochastic Model Applications 100%
• Meta-learning 100%
• Stochastic First-order Methods 100%
• Gradient Estimator 66%
• Sample Complexity 66%
• Non-convex Objective 66%
• Numerical Experiments 33%
• Logistic Model 33%
• Logistic Regression 33%
• Structural Assumption 33%
• Compositional Structure 33%
• Nonsmooth 33%
• Causal Inference 33%
• Lower Bound Analysis 33%
• Accelerated Algorithm 33%
• Stochastic Gradient Algorithm 33%
• Strongly Convex 33%
• Weak Convexity 33%
• Bias-variance Tradeoff 33%
• Lipschitz Continuous Gradient 33%
• Invariant Learning 33%
• Model-agnostic Meta-learning 33%

Mathematics

• Stochastics 100%
• Conditionals 100%
• Numerical Experiment 14%
• Variance 14%
• Logistic Model 14%
• Logistic Regression 14%
• Causal Inference 14%