During the first part of the week, the participants will follow classes where they will learn from the senior researchers. The second part of the week will let the participants become actors of the program and they will be asked to contribute to a large development project dedicated to the EPS topic. During this phase, they will apply what they have learned previously directly on concrete problems and examples.
The organization is balanced between some 'learning' time and some social activities. Every day will be separated in 2 half day sessions with a lunch break in the middle and a nice dinner. The dinner will be the occasion of discovering the French cuisine and some specialties from Brittany.

Main lectures

The students will follow several main lectures on metaheuristics, on matheuristics, on routing problems and on web-services. During the main lectures on matheuristics, the students will learn some important information about mathematical programming (CPLEX/GUROBI/FICO-XPress), about a new mathematical optimization solver (LocalSolver), and how to embed these tools in a C++ programming code. Two practical sessions will be given by practitioners from Artelys (FICO Xpress Optimization Suite) and Innovation24 (LocalSolver).

Participants gain an understanding of the Xpress suite, the Mosel modeling and programming language, and Xpress Workbench. Through a series of projects illustrating applications of the LP and MIP methods, students learn to implement optimization models using Xpress Optimization Suite.

Team work

For the interactive workshop, the students will be divided in small teams that will be coached by an experienced researcher. Each team will develop an optimization method for a given optimization problem, using a different approach: e.g., one team may use LocalSolver in combination with other metaheuristics techniques, other teams will integrate Gurobi or CPLEX or FICO-XPress together in a matheuristic framework. All these strategies will be proposed by the teams and supervised by the experienced researcher. All these techniques will be implemented in a web-service platform that is ready for use at a large scale. Onyx mac snow leopard 10.6 8. The web-service platform is able to provide real-life instances of routing problems.

Social activities

• Wed. 20 April: Welcome party
  
• Sat. 23 April: Beach party (or free time to visit the region depending on the weather)
• Sun. 24 April: Excursion to Groix island (boat trip + picnic)
• Wed. 27 April: Gala dinner

The FICO Xpress optimizer is a commercial optimizationsolver for linear programming (LP), mixed integer linear programming (MILP), convex quadratic programming (QP), convex quadratically constrained quadratic programming (QCQP), second-order cone programming (SOCP) and their mixed integer counterparts.^[2] Xpress includes a general purpose non-linear solver, Xpress NonLinear, including a successive linear programming algorithm (SLP, first-order method), and Artelys Knitro (second-order methods).

Xpress was originally developed by Dash Optimization, and was acquired by FICO in 2008.^[3]Its initial authors were Bob Daniel and Robert Ashford. The first version of Xpress could only solve LPs; support for MIPs was added in 1986.Being released in 1983, Xpress was the first commercial LP and MIP solver running on PCs.^[4]In 1992, an Xpress version for parallel computing was published, which was extended to distributed computing five years later.^[5]Xpress was the first MIP solver to cross the billion decision variable threshold by introducing 64-bit indexing in 2010.^[6]Since 2014, Xpress features the first commercial implementation of a parallel dual simplex method.^[2]

Technology[edit]

Linear and quadratic programs can be solved via the primal simplex method, the dual simplex method or the barrier interior point method. All mixed integer programming variants are solved by a combination of the branch and bound method and the cutting-plane method. Infeasible problems can be analyzed via the IIS (irreducible infeasible subset) method. Xpress provides a built-in tuner for automatic tuning of control settings.^[1]Xpress includes its modelling language Xpress Mosel^[7] and the integrated development environment Xpress Workbench^[8].Mosel includes distributed computing features to solve multiple scenarios of an optimization problem in parallel. Uncertainty in the input data can be handled via robust optimization methods.^[9]

Xpress has a modeling module called BCL (Builder Component Library) that interfaces to the C, C++, Java programming languages, and to the .NET framework.^[10] Independent of BCL, there are Python and MATLAB interfaces. Next to Mosel, Xpress connects to other standard modeling languages, such as AIMMS, AMPL, and GAMS.

The FICO Xpress Executor^[11] executes and deploys Mosel models, using SOAP or REST interfaces. It can be used from external applications or from the FICO Decision Management Platform.

References[edit]

1. ^ ^ab'FICO Xpress Optimization'. Jan 17, 2019.
2. ^ ^abBerthold, T.; Farmer, J.; Heinz, S.; Perregaard, M. (15 Jun 2017). 'Parallelization of the FICO Xpress-Optimizer'. Optimization Methods and Software. 33 (3): 518–529. doi:10.1080/10556788.2017.1333612.
3. ^'Dash Optimization acquired by FICO' Jan 22, 2008.
4. ^Ashford, R. (Feb 2007). 'Mixed integer programming: A historical perspective with Xpress-MP'. Annals of Operations Research. 149 (1): 5–17. doi:10.1007/s10479-006-0092-x.
5. ^Laundy, R. (1999). Implementation of Parallel Branch-and-bound Algorithms in XPRESS-MP. Operational Research in Industry. pp. 25–41. doi:10.1057/9780230372924_2. ISBN9780230372924.
6. ^O. Bastert (2011). FICO Xpress Optimization Suite(PDF) (Report). Retrieved Jan 23, 2019.
7. ^Guéret, Christelle; Prins, Christian; Sevaux, Marc (2002). Applications of Optimization with Xpress-MP. ISBN9780954350307.
8. ^'FICO Xpress Workbench'. Nov 12, 2017.
9. ^P. Belotti (2014). Robust Optimization with Xpress(PDF) (Report). Retrieved Oct 28, 2018.
10. ^'BCL Reference Manual' Nov 13, 2018.
11. ^'FICO Xpress Executor' Nov 13, 2018.

External links[edit]