Explore the cutting-edge field of Combinatorial Optimization Augmented Machine Learning (COAML) in this insightful DS4DM Coffee Talk presented by Maximilian Schiffer from the TUM School of Management at Technical University of Munich. Delve into the innovative fusion of machine learning and operations research techniques designed to tackle complex, contextual data-driven problems involving both uncertainty and combinatorics. Discover how COAML embeds combinatorial optimization layers into neural networks and utilizes decision-aware learning techniques to optimize industrial processes. Gain a comprehensive overview of the underlying paradigm, algorithmic pipelines, and foundations through selected application cases. Learn about the effectiveness of COAML in addressing contextual and dynamic stochastic optimization problems, including its award-winning performance in the 2022 EUROMeetsNeurIPS dynamic vehicle routing challenge.
Combinatorial Optimization Augmented Machine Learning for Contextual Multi-Stage Problems
GERAD Research Center via YouTube
Overview
Syllabus
Combinatorial optimization augmented machine learning for contextual multi-stage problems
Taught by
GERAD Research Center
Related Courses
-
Distributed Optimization and Machine Learning
-
Optimization problems and algorithms
-
Deep Learning for Combinatorial Optimization - Count Your Flops & Make Your Flops Count
-
Guiding Metaheuristics Through Machine Learning Predictions for Dynamic Autonomous Ridesharing
-
Approximation Algorithms for Stochastic Minimum Norm Combinatorial Optimization
-
Do Quantum Computers Have Applications in Machine Learning and Combinatorial Optimization
