How to integrate machine-learning probabilistic output in integer linear programming: A case for RSA

Abstract

We propose a methodology to integrate probabilistic outputs from machine learning models into integer linear programming formulations, demonstrated on the routing and spectrum assignment problem in optical networks.

Machine LearningOptical NetworksInteger Programming

BibTeX

@inproceedings{salani2019integrate,
  title        = {How to integrate machine-learning probabilistic output in integer linear programming: A case for RSA},
  author       = {Salani, Matteo and Rottondi, Cristina and Tornatore, Massimo},
  booktitle    = {45th European Conference on Optical Communication (ECOC 2019)},
  pages        = {1--4},
  year         = {2019},
  organization = {IET}
}