Inicio

OPTECO - research group from University of Málaga

Project from Spanish Ministry of Economy and Competitiveness

Title: NEW APPROACHES IN MULTIOBJECTIVE OPTIMIZATION THROUGH ALGORITHM-DECISION MAKER INTERACTION. APPLICATIONS TO ACADEMIC PERFORMANCE AND CHILD POVERTY

Reference: PID2020-115429GB-I00

Summary:

Among the wide variety of existing methodologies in multiobjective optimization (research field of Operations Research), this project aims to deepen the study of evolutionary multiobjective optimization algorithms in combination with interactive methods (widely used in multicriteria decision making). Currently, this area is identified among the experts in the evolutionary community as one of the most promising lines of research. In our case, we propose to develop new schemes that combine both methodologies, while establishing mechanisms that improve the incorporation of the preferences of the decision-maker (DM), so that situations of anchoring in the resolution process do not occur and the process ends once the DM reaches the most desired solution. 

Regarding evolutionary algorithms for multiobjective optimization, there is a deficit of metrics or indicators oriented to evaluate the solution capacity of algorithms and how good the solutions generated are according to the DM’s preferences. In this line, another of the objectives of this project focuses on the development of evaluation metrics for interactive multiobjective optimization evolutionary algorithms, establishing a comparative framework. 

On the other hand, the uncertainty involved in the mathematical modelling of many real processes makes it essential to have mathematical techniques capable of reflecting this uncertainty into the model. To cope with this, there exist intervalar programming techniques, very useful in econometric models, where we intend to study new formulations to extend the concept of achievement scalarizing function to intervalar programming. The idea is to be able to generate necessarily efficient solutions and, in case they do not exist, to obtain sufficiently efficient solutions for intervalar multiobjective problems. 

The multiobjective optimization methodologies proposed will be applied to two highly relevant socio-economic problems for the present and the future of the Spanish society: the improvement of academic performance of adolescents and the fight against child poverty. In policy terms, all governments assume the importance of education and of having children enjoying high levels of well-being, given the individual and collective implications of these two aspects for the economic and social development of the country. This project aims, on the one hand, to analyse in depth the connections between certain psychological, cognitive, social and physical aspects of the students’ wellbeing in Spain and their academic performance, based on PISA data. On the other hand, in relation to the well-being of children and adolescents, taking as a basis the European Union Statistics on Income and Living Conditions (EU-SILC), we propose to examine the main socio-economic factors determining the high levels of child poverty in Spain from a multidimensional perspective. For both applications, we propose models and formulations that reflect the reality analysed by means of multiobjective intervalar programming techniques, and our purpose is also to apply the interactive evolutionary algorithms developed in this project for solving them. The conclusions drawn from our analyses will allow progress to be made in the evaluation and design of effective measures, in terms of both educational policy and other public policies, which will make it possible to promote economic and social progress.

 

Project from Horizon Europe Framework Programme

Title: ENHANCED SAFE AND SUSTAINABLE COATINGS FOR SUPPORTING THE PLANET

Reference: PROPLANET

Call: HORIZON-CL4-2022-RESILIENCE-01-23

Summary:
PROPLANET addresses novel coating materials solutions, tackling the problem from a sustainable-business perspective, enabling overcoming the barrier for environmental protection, safety, chemical improvements, and circular value chains. The main goal of PROPLANET is to design and optimise 3 innovative coatings for industrial sectors: textile, food-packaging, and glass.
1) Crosslinked biopolymer oil/wax microcapsules in polysaccharide matrix (hydrophobic, olephobic bio-based coating)
2) Hybrid siloxane biobased coating (non-stick and anticorrosion protection bio-based hybrid coating)
3) Hybrid siloxane coating (anti-soiling (AS) anti-reflecting (AR) hybrid coating). The value chain of each product is optimised from raw material source to End Of Life of products, ensuring the circular economy. All coatings are designed based on Safety and Sustainability by design (SSbD) concepts and optimised with mathematical computational tools such as first-principles-based models, in silico models, environmental fate models, and sustainability assessment.

On top of that, replication activities aided by a powerful PROPLANET Replication tool based on data-driven algorithms and multiobjective optimization (MO) will promote the integration of the novel coatings on different applications, supporting their route to market, operating at different conditions, and following an Eco-design thinking. A well-balanced consortium, covering all actors in the , formed by end-users, technology solution providers and research organizations, ensures the successful achievement of objectives, which will allow a wide-spreading replication strategy towards efficient and safe designs for new coatings.