This paper introduces a new approach to enhance learning in adjustment processes by using a suppo... more This paper introduces a new approach to enhance learning in adjustment processes by using a support vector machine (SVM) algorithm as discriminant function jointly with an action generator module. The method trains a SVM with state-action patterns and uses trained SVM to select an appropriate action given a certain state in order to reach the target state. The system incorporates a simulated annealing technique to increase the exploration capacity and improve the ability to avoid local minima. The methodology has ...
Resumen: En el articulo que se presenta se utilizan tecnicas conexionistas de aprendizaje para re... more Resumen: En el articulo que se presenta se utilizan tecnicas conexionistas de aprendizaje para reproducir el proceso de evaluacion en la prediccion del riesgo de credito de las empresas. Como este proceso no se basa en informacion unicamente cuantitativa, sino que ademas tiene una fuerte dependencia del Conocimiento de los expertos, parece razonable la utilizacion de estos metodos. Las tecnicas de aprendizaje aplicadas son, en primer lugar, redes neuronales con funciones base radiales, y, en segundo lugar, maquinas de soporte ...
This paper lies within the domain of learning algorithms based on kernels of Support Vector Machi... more This paper lies within the domain of learning algorithms based on kernels of Support Vector Machines. A kernel is constructed over the discrete structure of absolute orders of magnitude spaces. This kernel is based on an explicit function, defined from the space of k-tuples of ...
This paper lies within the domain of supervised learning algorithms based on neural networks whos... more This paper lies within the domain of supervised learning algorithms based on neural networks whose architecture corresponds to radial basis functions. A methodology to use RBF when the descriptors of the patterns are given by means of their orders of magnitude is ...
Frontiers in Artificial Intelligence and Applications
Recent Advances in Artificial Intelligence Research and Development 111 J. Vitria el al.(Eds. 1 I... more Recent Advances in Artificial Intelligence Research and Development 111 J. Vitria el al.(Eds. 1 IOS Press, 2004 Qualitative Radial Basis Function Networks Applied to Financial Credit Risk Prediction Xari ROVIRA", Nuria AGELL", Monica SANCHEZ", Francesc PRATS+, ...
This paper presents the concepts underlying the design of an intelligent system based on the abso... more This paper presents the concepts underlying the design of an intelligent system based on the absolute orders of magnitude model to define qualitative operators able to dea l with a set of variables defined in different reference scale s and with different influence ...
This paper lies within the domain of supervised discretization methods. The methodology aims at i... more This paper lies within the domain of supervised discretization methods. The methodology aims at identifying relevant interactions between input and output variables. A new supervised discretization algorithm that takes into account the qualitative or- dinal structure of the output variable is proposed. Most existing supervised discretization methods are designed for pattern recognition problems and do not take into account this
Resumen Los algoritmos de aprendizaje basados en Funciones Núcleo, particularmente las Máquinas d... more Resumen Los algoritmos de aprendizaje basados en Funciones Núcleo, particularmente las Máquinas de Soporte Vectorial (MSV), han proporcionado buenos resultados en problemas de clasificación con patrones de entrada no separables linealmente. El uso de las Funciones Núcleo permite aplicar estos algoritmos de inferencia incluso sobre información proveniente de un conjunto sin estructura de espacio euclideo. Al considerar una Función Núcleo, los datos se proyectan de forma implícita sobre un nuevo espacio de ...
This paper lies within the domain of supervised discretization methods. The methodology aims at i... more This paper lies within the domain of supervised discretization methods. The methodology aims at identifying relevant interactions between input and output variables. A new supervised discretization algorithm that takes into account the qualitative ordinal structure of the output variable is ...
Abstract At the beginning of this new century, a growing concern for Ecology has led to the vital... more Abstract At the beginning of this new century, a growing concern for Ecology has led to the vital need for sustainability in society, the main reason being to achieve a high quality of life. This explains why, in the meeting of European municipalities in Hanover in February 2000, where the sustainability measurement indicators were agreed, the satisfaction of citizens with their local community was chosen as the first indicator. The social indicator movement began in the 1950s with objective health, education level, per capita income and life ...
This paper introduces a new approach to enhance learning in adjustment processes by using a suppo... more This paper introduces a new approach to enhance learning in adjustment processes by using a support vector machine (SVM) algorithm as discriminant function jointly with an action generator module. The method trains a SVM with state-action patterns and uses trained SVM to select an appropriate action given a certain state in order to reach the target state. The system incorporates a simulated annealing technique to increase the exploration capacity and improve the ability to avoid local minima. The methodology has ...
Resumen: En el articulo que se presenta se utilizan tecnicas conexionistas de aprendizaje para re... more Resumen: En el articulo que se presenta se utilizan tecnicas conexionistas de aprendizaje para reproducir el proceso de evaluacion en la prediccion del riesgo de credito de las empresas. Como este proceso no se basa en informacion unicamente cuantitativa, sino que ademas tiene una fuerte dependencia del Conocimiento de los expertos, parece razonable la utilizacion de estos metodos. Las tecnicas de aprendizaje aplicadas son, en primer lugar, redes neuronales con funciones base radiales, y, en segundo lugar, maquinas de soporte ...
This paper lies within the domain of learning algorithms based on kernels of Support Vector Machi... more This paper lies within the domain of learning algorithms based on kernels of Support Vector Machines. A kernel is constructed over the discrete structure of absolute orders of magnitude spaces. This kernel is based on an explicit function, defined from the space of k-tuples of ...
This paper lies within the domain of supervised learning algorithms based on neural networks whos... more This paper lies within the domain of supervised learning algorithms based on neural networks whose architecture corresponds to radial basis functions. A methodology to use RBF when the descriptors of the patterns are given by means of their orders of magnitude is ...
Frontiers in Artificial Intelligence and Applications
Recent Advances in Artificial Intelligence Research and Development 111 J. Vitria el al.(Eds. 1 I... more Recent Advances in Artificial Intelligence Research and Development 111 J. Vitria el al.(Eds. 1 IOS Press, 2004 Qualitative Radial Basis Function Networks Applied to Financial Credit Risk Prediction Xari ROVIRA", Nuria AGELL", Monica SANCHEZ", Francesc PRATS+, ...
This paper presents the concepts underlying the design of an intelligent system based on the abso... more This paper presents the concepts underlying the design of an intelligent system based on the absolute orders of magnitude model to define qualitative operators able to dea l with a set of variables defined in different reference scale s and with different influence ...
This paper lies within the domain of supervised discretization methods. The methodology aims at i... more This paper lies within the domain of supervised discretization methods. The methodology aims at identifying relevant interactions between input and output variables. A new supervised discretization algorithm that takes into account the qualitative or- dinal structure of the output variable is proposed. Most existing supervised discretization methods are designed for pattern recognition problems and do not take into account this
Resumen Los algoritmos de aprendizaje basados en Funciones Núcleo, particularmente las Máquinas d... more Resumen Los algoritmos de aprendizaje basados en Funciones Núcleo, particularmente las Máquinas de Soporte Vectorial (MSV), han proporcionado buenos resultados en problemas de clasificación con patrones de entrada no separables linealmente. El uso de las Funciones Núcleo permite aplicar estos algoritmos de inferencia incluso sobre información proveniente de un conjunto sin estructura de espacio euclideo. Al considerar una Función Núcleo, los datos se proyectan de forma implícita sobre un nuevo espacio de ...
This paper lies within the domain of supervised discretization methods. The methodology aims at i... more This paper lies within the domain of supervised discretization methods. The methodology aims at identifying relevant interactions between input and output variables. A new supervised discretization algorithm that takes into account the qualitative ordinal structure of the output variable is ...
Abstract At the beginning of this new century, a growing concern for Ecology has led to the vital... more Abstract At the beginning of this new century, a growing concern for Ecology has led to the vital need for sustainability in society, the main reason being to achieve a high quality of life. This explains why, in the meeting of European municipalities in Hanover in February 2000, where the sustainability measurement indicators were agreed, the satisfaction of citizens with their local community was chosen as the first indicator. The social indicator movement began in the 1950s with objective health, education level, per capita income and life ...
Uploads
Papers by Nuria Agell