Karina GIBERT : “The potentials of intelligent decision support systems in ecology”

vendredi 16 octobre 2015
par  Romain DAVID

Dep. Statistics and Operations Research
Knowledge Engineering and Machine Learning group
Universitat Politècnica de Catalunya, Barcelona, SPAIN
karina.gibert@upc.edu, www.eio.upc.edu/homepages/karina

New technologies made possible to get and store information of many different
types from any echological phenomenon. Currently, availability of data is not any more a limitation to understand reality. However, making a proper data exploitation to identify relevant decisional knowledge from it is still an open problem due to several reasons. On the one hand, the complexity of echological problems requires new data mining methods that can deal with high degree interactions, evolving along time and space. Also, data quality, as management of Big
Data and distributed information systems are relevant issues in these field. On the other hand, even in the cases where data quality is guaranteed, even where intensive data mining is possible, weven where rich knowledge is extracted from data, it is still difficult to guarantee a direct impact of data mining to effective decision-making levels. Intelligent Decision Support Systems (IDSS) provide a framework to conduct the flow between data recording, proper dataexploitation,
contextualization of data mining results, interpretation-support and results
communication to stakeholders in such a way that decision-making can be really supported by objective information obtained from data in a fluent cycle. Multi-perspective analysis and integral approach are required in most of the cases, to get reliable knowledge for supporting real decisions. In this work, the characteristics of echological data that cause poor performances in classical data mining will be discussed, the importance of data pre-processing steps will be analysed, the potentials of advanced data mining methods for different types of
echological applications will be introduced, and the potentials of IDSS to provide support to different kind of echological decisions will be explored.

Fayyad U, et al 1996 : From Data Mining to Knowledge Discovery : An overview. In Advances in Knowledge
Discovery and Data Mining. AAAI/MIT Press

Gibert, K, Sánchez-Marrè M, Rodríguez-Roda, I (2006) GESCONDA : An intelligent data analysis system for knowledge discovery and management in environemntal databases. Environmental modelling and software 21(1):115-120

K. Gibert, M. Sànchez-Marrè (2015) Improving ontological knowledge with reinforcement in recommending the data mining method for real problems. In Proceedings of Conferencia de la Asociación Española para la
Inteligencia Artificial (CAEPIA) 2015 (in press), Albacete, nov 2015

[Little 1970] J.D. Little. Models and Managers : The Concept of a Decision Calculus. Management Science
16(8):B466-85, 1970.

Misyak J, Giupponi C, Rosato P (2005) Towards the development of a decision support system for water resource
management. Environmental Modelling and Software 20(2):203-214
[Oprea 2005]

Pallottino, Stefano, Giovanni M. Sechi, and Paola Zuddas. "A DSS for water resources management under uncertainty by scenario analysis." Environmental Modelling & Software 20.8
(2005) : 1031-1042.

[Poch et al., 2004] Poch M, Comas J, R-Roda I, Sànchez-Marrè M & Cortés U. (2004) Designing and Building Real Environmental
Decision Support Systems Environmental Modelling & Software 19(9):857-873.

Rajasekaram V, Nandalal KDW (2005) Decision Support system for Reservoir Water management conflict resolution. Journal of water resources planning and management 131(3)410-









Aucun évènement à venir les 6 prochains mois