Machine Learning for Combinatorial Optimization: a Methodological Tour d'Horizon
This is an incremental survey that addresses the challenge of improving heuristic decisions in combinatorial optimization for researchers and practitioners in machine learning and operations research.
The paper surveys recent attempts to use machine learning for solving combinatorial optimization problems, advocating for deeper integration and proposing a methodology that treats optimization problems as data points to identify relevant distributions for learning.
This paper surveys the recent attempts, both from the machine learning and operations research communities, at leveraging machine learning to solve combinatorial optimization problems. Given the hard nature of these problems, state-of-the-art algorithms rely on handcrafted heuristics for making decisions that are otherwise too expensive to compute or mathematically not well defined. Thus, machine learning looks like a natural candidate to make such decisions in a more principled and optimized way. We advocate for pushing further the integration of machine learning and combinatorial optimization and detail a methodology to do so. A main point of the paper is seeing generic optimization problems as data points and inquiring what is the relevant distribution of problems to use for learning on a given task.