Deep neural networks and mixed integer linear optimization

• Published in: Constraints : an international journal Vol. 23; no. 3; pp. 296 - 309
• Main Authors: Fischetti, Matteo, Jo, Jason
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.07.2018; Springer Nature B.V
• Subjects: Artificial Intelligence; Artificial neural networks; Computation; Computer Science; Continuity (mathematics); Linear programming; Mixed integer; Neural networks; Operations Research; Operations Research/Decision Theory; Optimization; Topical Collection on Integration of Constraint Programming; Deep learning; Computational experiments; Deep neural networks; Mathematical optimization; Mixed-integer programming

Deep Neural Networks (DNNs) are very popular these days, and are the subject of a very intense investigation. A DNN is made up of layers of internal units (or neurons), each of which computes an affine combination of the output of the units in the previous layer, applies a nonlinear operator, and ou...