Modeling and Optimization Trade-off in Meta-learning

2020/10/24 by Katelyn Gao, Gao, Katelyn, Ozan Şener +1 · 1 citation
Computer Science · #Adversarial Robustness in Machine Learning #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #FOS: Mathematics #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Machine Learning and Data Classification #Optimization and Control (math.OC)

paper · pdf · doi:10.48550/arxiv.2010.12916

openalex publication_date 2020/10/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

By searching for shared inductive biases across tasks, meta-learning promises to accelerate learning on novel tasks, but with the cost of solving a complex bilevel optimization problem. We introduce and rigorously define the trade-off between accurate modeling and optimization ease in meta-learning. At one end, classic meta-learning algorithms account for the structure of meta-learning but solve a complex optimization problem, while at the other end domain randomized search (otherwise known as joint training) ignores the structure of meta-learning and solves a single level optimization problem. Taking MAML as the representative meta-learning algorithm, we theoretically characterize the trade-off for general non-convex risk functions as well as linear regression, for which we are able to provide explicit bounds on the errors associated with modeling and optimization. We also empirically study this trade-off for meta-reinforcement learning benchmarks.

Citations

• Trust Region Policy Optimization
• RL²: Fast Reinforcement Learning via Slow Reinforcement Learning
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• On First-Order Meta-Learning Algorithms
• Bilevel Programming for Hyperparameter Optimization and Meta-Learning
• Assessing Generalization in Deep Reinforcement Learning
• Meta-Learning with Implicit Gradients
• Rapid Learning or Feature Reuse? Towards Understanding the Effectiveness of MAML
• Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning
• SimpleShot: Revisiting Nearest-Neighbor Classification for Few-Shot Learning
• Rethinking Few-Shot Image Classification: a Good Embedding Is All You Need?
• On the Iteration Complexity of Hypergradient Computation

Cited by

• A Single-Loop Gradient Algorithm for Pessimistic Bilevel Optimization via Smooth Approximation

Related

• On the Global Optimality of Model-Agnostic Meta-Learning
• Weighted Meta-Learning
• Adaptive Gradient-Based Meta-Learning Methods
• Meta-learning with negative learning rates
• Task-Robust Model-Agnostic Meta-Learning