Researchers develop a mathematical model to facilitate making decisions
July 16, 2013
The model aims at representing mathematically the imprecise expressions used in human language. Finances, artificial intelligence or computer science are some of the fields in which this research may be applied.
What is the best decision when there are several different alternatives and the information we have about them is not concrete or certain? Mathematics are a precise tool to define models that help you decide which is the most rational decision in the processes of decision-making.
The research group on Modelization of Uncertainty and Imprecision in Decision Theory (UNIMODE) is finishing a mathematical model, which includes elements from the so-called fuzzy lógic, to try to make mathematics represent and translate the imprecisions of human language. It is an innovate approach that goes beyond classic mathematics, since it tries to discover the mathematical expression of the quantifiers of our language.
The researchers from UNIMODE have been working to make those mathematical models get closer to the way people reason and to make them represent imprecise concepts, such as "a lot", "a few" or "much", etc. The Ministry of Economics and Competitivity funds, through the National Plan of Science, the project titled Modelization of Uncertainty and Imprecision in the Making of Decisions, whose results may have an important impact in fields like economics and finances, computer science, robotic development or the advance of artificial intelligence.
The UNIMODE research group collaborates with enterprises of the private sector to make an attempt at modelizing the problems of human reliability, which cause between 70 and 90 per cent of labor accidents
The results of the project are still being finalized, but some of them already have a practical application. The final goal of the researchers is to develop a mathematical theory on the ordenation of alternatives linked to a random experiment under an approximate reasoning. This would facilitate the process to the person making the decision, allowing them to choose between different alternatives in a more real context. This is why artificial intelligence would be one of the fields in which these models would be most valuable.
Application in the enterprise
In order to test the theoretical results, the researchers of the UNIMODE are working in collaboration with teachers of other departments of the University of Oviedo, applying part of their study to prevent the attacks on users' privacy that can be found online.
The application of these mathematical models in processes of production or decision in the enterprise is key for optimizing its results. The researchers of the UNIMODE collaborate, for example, with some Asturian enterprises to try to formalize the problems of human reliability, which cause around 70 and 90 per cent of labor accidents. Creating a mathematical tool that represents exactly those human reliability problems may be an strategic step forward in the field of healthcare and labor security.
The impact of the research conducted by UNIMODE has been very relevant. More than 30 international scientific journals have published articles on the most significant advances of the project and their researchers have authored more than 20 chapters in specialized books.
Moreover, UNIMODE has a vast network of collaborators in other prestigious universities and research centers. The Asturian researchers collaborate with groups in the University of Ghent (Belgium), the Universities of Pavia and Trieste (Italy), the University of Mateja Bela (Slovakia), the University of Novi Sad (Serbia), the Dalle Molle Institute of Studies on Artificial Intelligence (Switzerland), the University of Valladolid, the University of Granada, the Public University of Navarre, the University of Salamanca or the University of Málaga.
What is a risky decision?
The starting point of this study by the researchers of UNIMODE are those situations in real life in which it is not possible to fully-accurately determine the probability of each of the alternatives taking place. These are the cases in which we find ourselves with a risky decision, or an uncertain decision. For example, an investor who has to distribute their resources among different alternatives, or a subject that must decide whether they buy or not a lottery ticket. According to the economic terminology, these consequences in which uncertainty is a factor are called perspectives or lotteries.
Research Group on Modelization of Uncertainty and Imprecision in Decision Theory (UNIMODE)
- Susana Montes (Coordinator)
- Susana Díaz
- Tania Iglesias
- Davide Martinetti
- Enrique Miranda
- Ignacio Montes
- Ekra Sanggala