LMCA　 2020 - Learning Meets Combinatorial Algorithms Workshop at NeurIPS 2020

Why this Workshop?
Machine learning algorithms have been shown to generalize poorly on combinatorially demanding tasks. Recent research has demonstrated that merging combinatorial optimization with machine learning methods enables solving problems that require non-trivial combinatorial generalization beyond pattern matching. In this spirit, this workshop aims to bring the communities (machine learning and combinatorial optimization, operations research) together in order to motivate further research at the intersection. This involves:
Machine learning approaches aimed at improving combinatorial algorithms/solvers.
Machine learning techniques to directly learn solvers for combinatorial problems.
Hybrid architectures; pipelines containing both algorithmic/combinatorial and standard NN building blocks.
Applications of the above.
Call for Papers
We will be accepting abstracts (4 pages excluding acknowledgement section, references and appendix) that fit the theme of the workshop. Please use the standard NeurIPS template for submitting the abstracts, the submission may optionally contain an appendix, no broader impact section is required. We do not accept re-submissions of accepted papers to conferences (including NeurIPS). Authors of accepted abstracts are expected to provide a short video (5 min) describing their work and to participate in a short poster session on the day of the workshop.