Research Group in Intelligent Systems
N�ria Maci� » Projects

Projects



Current Projects

  • INTEGRIS: Intelligent Electrical Grid Sensor Communications, 2010-2012.
  • KEEL III: Knowledge Discovery based on Evolutionary Learning: Current Trends and New Challenges (KEEL-CTNC), 2009-2011.
  • SGR2009 : Grup de recerca consolidat per la Generalitat de Catalunya (Consolidated Research Group, accredited by Generalitat de Catalunya), 2002-2004, 2005-2008, 2009-2011.
  • MINDATOS: Spanish Network on Data Mining and Machine Learning (Red Española de Minerí­a de Datos y Aprendizaje Automático), 2010-present.


Previous Projects

  • GAD: Active Demand Management, 2006-2010.
  • MID-CBR-GRSI: A Unified Framework for the Development of Case-Based Reasoning Systems, Un Marco Integrador para el Desarrollo de Sistemas de Razonamiento Basado en Casos, 2006-2009.
  • KEEL II: Knowledge Extraction Based on Evolutionary Learning: Evolutionary Rule-Based Systems. Data Mining Applications. Data Complexity and Design of Experiments, 2005-2008.
  • IA-Simusold: Artificial Intelligence-Based Training for a Virtual Welding Simulator, 2005.
  • Active Career (I,II), 2004-2005.
  • Intelligent Robots, 2005.
  • KEEL: Knowledge Extraction based on Evolutionary Learning: Rule Learning based on Michigan and Pittsburgh Approaches, Case-based Reasoning, and Data Mining, 2002-2005.
  • HRIMAC: AI tool for the Content-based Retrieval of Mammography Images applied to the Diagnosis of Breast Cancer, 2003-2005.
  • XIA2003: Participation in the Catalan Artificial Intelligence Network, 1999-2005.
  • FIS2000: Prediction of Breast Cancer risk based on Mamographies analysed with Computer Vision and Machine Learning Techniques (FIS 00/0033-02), 2000-2003.
  • Automatic classification of long bone fractures from X-Ray images using the Müller’s CCF system, Maurice E. Müller Fundation, 1999-2001.
  • Computer insfrastructure, 1997-2001.
  • AI-elearning-networks: Analysis and Implementation of AI tools for e-learning and ATM networks (CICYT/TEL98-0408), 1998-2000.
  • Automatic prediction of forest fire risk (Funitec, CTFC), 1999.
  • Generation of acustic models of streets with AI tools (Funitec).



[top]

INTEGRIS

Project Intelligent Electrical Grid Sensor Communications, (2010-2012)

Main Researcher

Dr. Josep Maria Selga

Reference

FP7-ICT-ENERGY-2009-1

Description

INTEGRIS project proposes the development of a novel and flexible ICT infrastructure based on a hybrid Power Line Communication-wireless integrated communications system able to completely and efficiently fulfil the communications requirements foreseen for the Smart Electricity Networks of the future.
This includes encompassing applications such as monitoring, operation, customer integration, demand side management, voltage control, quality of service control, control of Distributed Energy Resources and asset management and can enable a variety of improved power system operations, some of which are to be implemented in field trials that must proof the validity of the developed ICT infrastructure.
Focus is on interoperability of the PLC, Wireless Sensor Network and Radio Frequency Identification, technologies that together are able to achieve the indicated goal with reasonable cost. The system will require an adequate management system that is also an objective of the project. Such system will be based on beyond the state-of-the-art cognitive techniques to provide the system with the adequate flexibility, scalability, availability, security, enhanced system life-time and self-healing properties as is necessary in complex and dynamic systems.
A further objective is to research on the limits and benefits of distributing smart grid applications in the newly designed INTEGRIS system. This will have an impact on the availability of those applications and influence the developed devices and platforms since they will require a certain level of storage and computing capabilities. The final aim of the INTEGRIS project is to provide an ICT system that enables the improvement of the performance of the electricity distribution grid in agreement with the impact foreseen in the work program.
The INTEGRIS project is a cross thematic research approach integrating knowledge and partners from ICT and Energy fields and aims to create and consolidate such a cross-thematic team.

Link

Not available

Members

Endesa Network Factory SL., Indra Sistemas SA, Diseño de Sistemas en Silicio SA, FUNITEC - La Salle - Universitat Ramon Llull, Schneider Electric Industries SAS, Current Technologies International GMBH, iLight, A2A Reti Elettriche SPA, Tampereen Teknillinen Yliopisto

Funded by

European Comission.

Period

2010-2012



[top]

KEEL III

Project KEEL III: Knowledge Discovery based on Evolutionary Learning: Current Trends and New Challenges (2009-2011)

Main Researcher

Dr. Ester Bernadó-Mansilla

Reference

TIN2008-06681-C06-05

Description

The project KEEL-CTNC is the continuation of the Project KEEL II (http://www.keel.es), which is under development for the period January 2009 – December 2011. KEELCTNC is focused on the knowledge extraction from data using evolutionary algorithms, and it is comprised of four global objectives.

1) The first one is to continue with the development of the KEEL software tool (http://www.keel.es) that integrates the construction and use of specific modules collecting the state of the art of algorithms in specific topics of knowledge discovery with evolutionary learning such as genetic fuzzy systems, algorithms for learning from low quality data, imbalanced data set learning algorithms, …

2) The second objective is to continue with the development of evolutionary learning models and/or their improvement and adaptation to specific contexts associated to the current trends on knowledge extraction based on evolutionary learning: genetic fuzzy systems, evolutionary neural networks, learning classifier systems, evolutionary learning from low quality data, evolutionary learning algorithms for imbalanced data sets, multi-instance learning, data complexity evaluation for evolutionary learning, …

3) The development of studies on new challenges in knowledge extraction based on evolutionary learning, fixing our attention on a challenge problem as “Scaling Up for High Dimensional Data and Large Data sets”, the scaling up and high dimensional problems.

4) The fourth and last one, the characterization of specific real problems and the applicability of our evolutionary learning algorithms: marketing, chemical agroalimentary problems, mobile robotics, educational data mining (e-learning) and web mining.

For more details, visit the KEEL home page.

Links

KEEL Home Page

Paper: J. Alcalá-Fdez, L. Sánchez, S. Garcí­a, M.J. del Jesus, S. Ventura, J.M. Garrell, J. Otero, C. Romero, J. Bacardit, V.M. Rivas, J.C. Fernández, F. Herrera. KEEL: A Software Tool to Assess Evolutionary Algorithms to Data Mining Problems. Soft Computing 13:3 (2009) 307-318. PDF

Members

Universidad de Granada, Universidad de Jaén, Universidad de Córdoba, Universidad de Oviedo, Universitat Ramon Llull

Funded by

Ministerio de Ciencia y Tecnología, Fondo Europeo de Desarrollo Regional (FEDER).

Period

2009-2011



[top]

MID-CBR-GRSI

Project A Unified Framework for the Development of Case-Based Reasoning Systems, Un Marco Integrador para el Desarrollo de Sistemas de Razonamiento Basado en Casos, (2006-2009)

Main Researcher

Dra. Elisabet Golobardes i Ribé

Reference

TIN2006-15140-C03-03

Description

Case-based reasoning (CBR) combines in an effective manner both learning from examples and usage of domain knowledge. Techniques for case retrieval and reuse should not be studied in an isolated manner; instead they should be designed and evaluated in a framework that integrates different types of CBR systems as proposed in this project.

The main objectives of the project can be summarized as follows:

    new ways to use techniques of soft computing for CBR,
    techniques for case reuse of a declarative and generic nature,
    techniques for case retrieval in knowledge-intensive CBR systems,
    integrating ontologies both in CBR systems and retrieval and reuse techniques,
    maintenance techniques both for case bases and for CBR systems capable of dealing with issues arising from design, implementation, and deployment of industrial strength CBR systems,
    the empirical evaluation of the developed techniques by means of CBR prototypes implemented for several experimental domains, and
    developing component-based software platforms to support CBR systems development.

Link

Not available

Members

Instituto de Investigación de Inteligencia Artificial, Universidad Ramon Llull, Universidad Complutense de Madrid

Funded by

Ministerio de Ciencia y Tecnología, Fondo Europeo de Desarrollo Regional (FEDER).

Period

2006-2009



[top]

KEEL II

Project KEEL II: Knowledge Extraction Based on Evolutionary Learning: Evolutionary Rule-Based Systems. Data Mining Applications. Data Complexity and Design of Experiments (2005-2008)

Main Researcher

Dr. Josep M. Garrell i Guiu

Reference

TIN2005-08386-C05-04

Description

KEEL is a software tool to assess evolutionary algorithms for Data Mining problems including regression, classification, clustering, pattern mining and so on. It contains a big collection of classical knowledge extraction algorithms, preprocessing techniques (instance selection, feature selection, discretization, imputation methods for missing values, etc.), Computational Intelligence based learning algorithms, including evolutionary rule learning algorithms based on different approaches (Pittsburgh, Michigan and IRL, …), and hybrid models such as genetic fuzzy systems, evolutionary neural networks, etc. It allows us to perform a complete analysis of any learning model in comparison to existing ones, including a statistical test module for comparison. Moreover, KEEL has been designed with a double goal: research and educational. For more details, visit the KEEL home page.

Links

KEEL Home Page

Paper: J. Alcalá-Fdez, L. Sánchez, S. Garcí­a, M.J. del Jesus, S. Ventura, J.M. Garrell, J. Otero, C. Romero, J. Bacardit, V.M. Rivas, J.C. Fernández, F. Herrera. KEEL: A Software Tool to Assess Evolutionary Algorithms to Data Mining Problems. Soft Computing 13:3 (2009) 307-318. PDF

Members

Universidad de Granada, Universidad de Jaén, Universidad de Córdoba, Universidad de Oviedo, Universitat Ramon Llull

Funded by

Ministerio de Ciencia y Tecnología, Fondo Europeo de Desarrollo Regional (FEDER).

Period

2005-2008



[top]

GAD

Project GAD: Gestión Activa de la Demanda (Active Demand Management)
Description The GAD project is on the management of power demand. The project has three main objectives: a) characterization of the user’s profile, b) prediction of the power demand, c) design and development of action algorithms to re-balance the power load. Clustering techniques will be designed and applied on a test scenario.
Main Researcher: Dr. Josep M. Garrell Guiu
Reference CEN200710126
Link
Members GTD, Iberdrola, Red Eléctrica Española, Unión Fenosa, Siemens, Bosch, Fagor…
Funded by: Ministerio de Industria, Turismo y Comercio
Period 2006-February 2010



[top]

SGR 2009

Project Consolidated Research Group, accredited by Generalitat de Catalunya (Suport a Grups d’Investigació Consolidats de Catalunya)
Main Researcher: Dr. Ester Bernadó-Mansilla
Reference 2009 SGR-00183, 2005 SGR-00302, 2002 SGR-00155
Link DURSI
Funded by: Departament d’Universitats, Recerca i Societat de la Informació (DURSI) - Generalitat de Catalunya
Period 2009-2011, 2005-2008, 2002-2004 (former project)



[top]

MINDATOS

Project Participation in the Spanish Network of Data Mining and Machine Learning
Main Researcher: Dr. José Cristóbal Riquelme Santos
Reference TIN2010-09163-E
Link Main researcher’s home page
Funded by: Ministerio de Ciencia y Tecnología
Period 2010-present


[top]

IA-Simusold

Project
IA-Simusold: Artificial Intelligence-Based Training for a Virtual Welding Simulator
Description

The project consists in developing a system based on machine learning techniques which will be able to predict the quality of welds performed in a virtual welding simulator. The welding simulator was designed to train welders by means of a simulated environment similar to a real welding environment. The simulator, called Simusold, permits trainees to learn how to weld by means of a virtual environment. This reduces people’s training time of real weldings, which consequently reduces the economic cost and minimizes the high accident associated to real weldings of non-trained people. Simusold was developed in a previous project. In this project, we aim at integrating an intelligent system which evaluates automatically the ability of the user and the quality of the finished weld. The system is based on a knowledge representation that gathers the knowledge of welding experts and uses machine learning techniques to subsequent refine the model. The system is able to assess an estimation of the quality of the weld performed by the trainee and moreover, may give suggestions on how to improve the quality of welds in further simulated welding episodes.

Main Researcher Dra. Ester Bernadó Mansilla
Reference CIT-020600-2005-42
Link
Members Grup de Recerca en Sistemes Intel.ligents, Institut Català  Tècnic de la Soldadura, Departament de Tecnologies Audiovisuals (secció Multimèdia) d’Enginyeria i Arquitectura La Salle (URL)
Funded by Ministerio de Ciencia y Tecnología
Period 2005



[top]

Active Career I and II

Project Active Career II
Description

This is the second part of project Active Career. The aim of this project is to complete the prototype developed in Active Career I.
Active Career is a system, supported by artificial intelligence techniques, that allows to predict the evolution of job offers as well as the evolution of the profile of candidates. Also, the system will find the proper matchings between a given candidate and the job offers. The application will take information from several domains (web pages, company intranets, universities, etc).

Main Researcher Dra. Elisabet Golobardes (direcció IA)
Reference Company contract (part II), FIT-350100-2004-255 (part I)
Members Grup de Recerca en Sistemes Intel.ligents (GRSI), Transferència de Tecnologia La Salle (also Universidad de Navarra in part I)
Funded by Active Career (part II), Ministerio de Ciencia y Tecnologí­a (part I)
Period 1/06/2005 - 1/09/2006



[top]

Robots

Project
Robots
Description

This is to study reinforcement learning agents in real environments. We aim at investigating several learning algorithms and test them on real environments. The types of applications developed are: obstacle avoidance, map construction, orientation in the environment, etc. The aim of the project is to favor the participation of students.

Main Researcher Alvaro Garcí­a, Joan Camps, Dra. Ester Bernadó Mansilla
Reference 2005 PEIR 0059/17
Funded Ajut d’infrastuctura del Departament d’Universitats, Recerca i Societat de la Informació (DURSI), Generalitat de Catalunya
Period 2005



[top]

KEEL

Nom KEEL: Knowledge Extraction based on Evolutionary Learning: Rule Learning based on Michigan and Pittsburgh Approaches, Case-based Reasoning, and Data Mining, 2002-2005.
Description

This is part I of project KEELII. See KEEL II for the details of the current project.
The goal of KEEL part I was: to design and implement a prototype for knowledge extraction based on evolutionary learning. The project was developed by five Spanish universities. The goal of our research group was to integrate evolutionary rule-based systems, specifically those based on the Michigan and Pittsburgh approaches. Also, case-based reasoning and other techniques of data mining were included.

Main Researcher Dr. Josep M. Garrell i Guiu
Reference TIC2002-04036-C05-03
Link KEEL home page
Members Universidad de Granada, Universidad de Jaén, Universidad de Córdoba, Universidad de Oviedo, Universitat Ramon Llull
Funded by Ministerio de Ciencia y Tecnología, Fondo Europeo de Desarrollo Regional (FEDER).
Period 2002-2005



[top]

HRIMAC

Project HRIMAC: AI tool for the Content-based Retrieval of Mammography Images applied to the Diagnosis of Breast Cancer, 2003-2005.
Descripció El projecte forma part del projecte coordinat:

“Herramienta de recuperación de imágenes Mamogríficas por análisis de contenido (HRIMAC) para el asesoramiento en el diagnóstico del cáncer de mama ”

El resum d’aquest projecte coordinat és:

“El proyecto HRIMAC está concebido para funcionar como un sistema de recuperación de imágenes por contenido que permita acceder a una determinada tipologí­a de imágenes de mamografí­as digitales almacenadas en las diversas bases de datos públicas, a partir del contenido de una imagen ejemplo, siguiendo determinados criterios de afinidad. Así­, a partir de una imagen mamográfica sobre la cual se pretende emitir un diagnóstico, HRIMAC debe buscar en las diferentes bases de datos mamográficas las N mamografí­as más similares, de acuerdo con los criterios especificados en la búsqueda. De esta manera, cada búsqueda proporcionará un conjunto limitado de casos (mamografías digitales) con ciertas características (la forma de los clusters de microcalcificaciones, la presencia de determinadas lesiones espiculares, la forma de las masas, etc.) muy similares a la mamografí­a que se somete a estudio. El análisis de estos casos, ya patológicos, puede sin lugar a dudas ayudar al radiólogo a diagnosticar con más garantías de éxito, y aumentar de esta forma el grado de eficacia en la interpretación. El objetivo final del proyecto es desarrollar un prototipo de aplicación que funcione sobre plataforma web, y que dote al colectivo de radiólogos con información adicional relevante sobre el caso que se analiza, para así facilitar la emisión de un diagnóstico más fiable sobre el cáncer de mama. Para ello, es necesaria la combinación de técnicas de Visión por Computador (encaminadas a la extracción eficiente de características de las imágenes) con algoritmos de Aprendizaje Artificial (orientados a la selección de caracterí­sticas y a la asociación de casos). HRIMAC también podrá ser utilizado como una herramienta didáctica, que permita ejemplificar casos prototipos de las diferentes sintomatologías registradas en el análisis de imágenes mamográficas.”

Investigador principal: Dra. Elisabet Golobardes
Referància TIC2002-04160-C02-02
Enllaç Pàgina principal del projecte coordinat
Participants Universitat de Girona (UdG), Hospital Universitari Doctor Josep Trueta de Girona (HJT), Universitat Ramon Llull (URL)
Finançat per: Ministerio de Ciencia y Tecnologí­a, Fondo Europeo de Desarrollo Regional (FEDER).
Durada 2002-2005


[top]

XIA 2003

Nom Participació en la Xarxa Temática d’Intel·ligència Artificial
Descripció La missió fonamental de la Xarxa Temática en Intel·ligència Artificial (XIA) és la coordinació i l’aprofundiment del coneixement mutu dels equips de recerca. Les activitats de la XIA s’enfocaran envers la docència, la recerca i la transferència de tecnologia dins l’àmbit de la Intel·ligència Artificial a Catalunya.
Investigador principal: Dr. Josep Puyol-Gruart
Referència 2003XT00075
Enllaç Pàgina principal XIA
Participants Institut d’Investigació en Intel·ligència Artificial (CSIC), Enginyeria de Sistemes i Automàtica - Sistemes intel·ligents de control (Universitat de Girona), Grup d’Intel·ligència Artificial (Universitat de Lleida), Secció d’Intel·ligència Artificial (Universitat Politècnica de Catalunya), Grup de Recerca en Intel·ligència Artificial (Universitat Rovira i Virgili), Centre de Visió per Computador (Universitat Autònoma de Barcelona), Grup de Recerca en Sistemes Intel·ligents (Universitat Ramon Llull).
Finançat per: CIRIT
Durada 2003-?


[top]

FIS 2000

Nom Sistema de predicción del riesgo de cáncer de mama basado en técnicas de Visión por Computador y de Aprendizaje Artificial Automático
Descripció
Investigador principal: Dra. Elisabet Golobardes
Referència FIS 00/0033-02
Enllaç
Participants Universitat de Girona (UdG), Hospital Universitari Doctor Josep Trueta de Girona (HJT), Universitat Ramon Llull (URL)
Finançat per: Instituto de Salud Carlos III, Fondo de Investigación Sanitaria, Ministerio de Sanidad y Consumo
Durada 2000-2003


[top]

XIA 2001

Nom Participació en la Xarxa Temàtica d’Intel·ligència Artificial
Descripció La missió fonamental de la Xarxa Temàtica en Intel·ligència Artificial (XIA) és la coordinació i l’aprofundiment del coneixement mutu dels equips de recerca. Les activitats de la XIA s’enfocaran envers la docència, la recerca i la transferència de tecnologia dins l’àmbit de la Intel·ligència Artificial a Catalunya.
Investigador principal: Dr. Pere García
Referència 2000XT-00031
Enllaç
Participants
Finançat per: CIRIT/DURSI - Generalitat de Catalunya
Durada 2001-2003


[top]

XIA 1999

Nom Participació en la Xarxa Temàtica d’Intel·ligència Artificial
Descripció La missió fonamental de la Xarxa Temàtica en Intel·ligència Artificial (XIA) és la coordinació i l’aprofundiment del coneixement mutu dels equips de recerca. Les activitats de la XIA s’enfocaran envers la docència, la recerca i la transferència de tecnologia dins l’àmbit de la Intel·ligència Artificial a Catalunya.
Investigador principal: Dr. Vicenç Torra
Referència 1998XT000xx
Enllaç
Participants
Finançat per: CIRIT/DURSI - Generalitat de Catalunya
Durada 1999-2001


[top]

IA-Learning 1998

Nom Implementación y estudio de herramientas de inteligencia artificial aplicadas a plataformas de enseñanza abierta a distancia y a las redes ATM que las soportan
Descripció
Investigador principal: Dr. Josep M. Garrell
Referència CICYT/TEL98-0408
Enllaç
Participants Universitat de Girona, Enginyeria i Arquitectura La Salle (Universitat Ramon Llull), Universitat Oberta de Catalunya, GRN Serveis Telemàtics S.L., MIFAS (Minusvàlids Físics Associats, S.A.), Telefónica de España S.A., DIMAT S.A.
Finançat per: Comisión Interministerial de Ciencia y Tecnologí­a, Ministerio de Educación
Durada 1998-2000