Instituto de Matemáticas
Universidad de Sevilla

Sevilla, 20th October, 2017

Emilio Carrizosa Priego

Instituto de Matemáticas de la Universidad de Sevilla

He is Professor of Statistics and Operations Research, President of the Spanish Society of Statistics and Operations Research, former Editor-in-Chief of the OR journal TOP (Q3).

He works mainly on three overlapping topics:

  • Nonlinear Optimization and its application to Operations Research models.
  • Optimization in Data Analysis, addressing problems in Classification, and Dimensionality Reduction Techniques.
  • Applications of Nonlinear Programming to industrial problems.

In a) he has studied structural properties of selected nonlinear problems, identifying sets which reduce the search of an optimal solution to inspecting the points in a grid. A second strategy was to write the problem as a d.c. problem, i.e., involving functions expressed as difference of two convex functions. Finding an adequate d.c. representation has shown to be extremely useful in branch and bound algorithms. These two approaches only work properly for small-dimensional problems. For larger problems, heuristics are the unique choice. He has been working with P. Hansen and N. Mladenovic, creators of the off-the-shelf metaheuristic Variable Neighborhood Search, applied to some problems in Locational Analysis.

In b) he has sought solutions easier to interpret (more parsimonious) in classic problems in Data Analysis. The aim of selecting a few features or a few records has frequently led to nonlinear optimization problems in integer numbers, addressed with the techniques developed in a).

Lines a) and b) have been financed by national and regional research projects since 1996.

Line c) has being developed thanks to different R&D contracts, including those with Universal DX (data analysis in metabolomics), Abengoa Solar (new algorithms for the optimal design of solar plants), REPSOL (Global Optimization in inverse problems appearing in parameters estimation in metabolic networks), Taitus (optimal management of satellites), Junta de Andalucía (Optimization tools to forecast dry periods), and the EU-funded Project COST TD 1207, being one of the two Spanish members in its Management Committee.

He is the Principal Investigator of the ERA net (H2020) entitled “Optimal Heliostat Fields for Solar Tower Power Plants”, and of the 2016 Big Data-project entitled “COST-Sensitive classification. A mathematical optimization approach”, funded by the foundation of the bank BBVA.