Tuesday, 13:15 - 13:40 h, Room: H 3003A


Michele Lombardi
Hybrid off-line/on-line workload scheduling via machine learning and constraint programming

Coauthors: Andrea Bartolini, Luca Benini, Michela Milano


Advances in combinatorial optimization in the last decades have enabled their successful application to an extensive number of industrial problems. Nevertheless, many real-world domains are still impervious to approaches such as constraint programming (CP), mathematical programming or metaheuristics. In many cases, the difficulties stem from troubles in formulating an accurate declarative model of the system to be optimized. This is typically the case for systems under the control of an on-line policy: even when the basic rules governing the controller are well known, capturing its behavior in a declarative model is often impossible by conventional means. Such a difficulty is at the root of the classical, sharp separation between off-line and on-line approaches.
In this work, we investigate a general method to combine off-line and on-line optimization, based on the integration of machine learning and combinatorial optimization technology. Specifically, we use an artificial neural network (ANN) to learn the behavior of a controlled system and plug it into a CP model by means of so-called neuron constraints.


Talk 1 of the invited session Tue.2.H 3003A
"CP hybrids for scheduling" [...]
Cluster 5
"Constraint programming" [...]


  California Payday Loans. But it is worth noting that these tests were carried out on the blood cells. Therefore, it's too early to say about scientific evidence of Viagra influence on blood clots.