Objectifs

The main motivation today for the study of optimization is the improvement of industrial processes. But there are few areas of knowledge where its application does not extend to. It is intensively used in logistics, transportation, scheduling, resource management, finance ; but also biology, chemistry, physics, mathematics, economics, linguistics, philosophy, and more. Within engineering, electrical (INF569), mechanical (MAP562), and chemical are those on which it has the most important impact.

Optimization rests on a formal descriptive language called mathematical programming (MAP557, INF569, INF580), which is used to construct a formal model, called mathematical program (MP), of a given optimization problem. An MP is usually solved by an off-the-shelf piece of software called solver. A solver accepts all of the MP instances of a certain class and given enough time, outputs the solution. Very large-scale MPs that defy general-purpose solvers need more investigation, work and algorithmic creativity to be solved (INF580, INF550). It is sometimes necessary to resort to hybrid approaches involving randomized heuristics (INF561), artificial intelligence (CSC_52081_EP and machine learning (INF554, MAP569).