Meta Learning Control Variates for Variance Reduction with Limited Data
Offered By: Uncertainty in Artificial Intelligence via YouTube
Course Description
Overview
Explore an innovative approach to variance reduction in Monte Carlo estimators through a 25-minute oral presentation from the Uncertainty in Artificial Intelligence conference. Delve into the concept of Meta Learning Control Variates (Meta-CVs), a powerful technique designed to improve performance when computing multiple related integrals with limited samples. Discover how this method leverages similarities between integration tasks to achieve significant variance reduction, even in scenarios with very small sample sizes. Gain insights into the empirical assessment and theoretical analysis that establish the conditions for successful Meta-CVs implementation. Access the presentation slides to visualize key concepts and findings from this cutting-edge research in artificial intelligence and statistical computing.
Syllabus
UAI 2023 Oral Session 5: Meta Learning Control Variates Variance Reduction with Limited Data
Taught by
Uncertainty in Artificial Intelligence
Related Courses
Hiper-Memória & Hiper-AprendizagemUdemy Stanford CS330: Deep Multi-Task and Meta Learning
Stanford University via YouTube Stanford Seminar - The Next Generation of Robot Learning
Stanford University via YouTube Parameter Prediction for Unseen Deep Architectures - With First Author Boris Knyazev
Yannic Kilcher via YouTube Efficient and Modular Implicit Differentiation - Machine Learning Research Paper Explained
Yannic Kilcher via YouTube