Tratamiento, indexación y recuperación de información en grandes volúmenes de datos mediante técnicas de reconocimiento de formas: aplicaciones en descripción y clasificación de datos secuenciales
A change to follow the last tendencies is proposed according to recent relevant works published in international forums from the Pattern Recognition and Machine Learning domains. On the other hand, some extremely relevant applications are considered, both because of they appear as high-priority thematic lines in the National Program and because of their importance in other international initiatives, mainly in the European area.
From a technical viewpoint, the main objectives refer to the development and adaptation of techniques in which the groups have acquired enough skill, expertise and acknowledgement (efficient classification in metric spaces, feature and prototype selection, image classification, stochastic models), to situations and cases of practical interest where the common fact is the very large size of (all aspects of) the problems. In addition, we propose to get into other research lines of recent interest like ensembles of classifiers, information fusion, ongoing learning, adaptive classification and non conventional techniques of clustering and data organization.
Finally, the project also tries to partly address the research of the group towards application areas that today constitute a convergence of concepts, techniques and solutions from of close fields like machine learning and artificial intelligence, information retrieval, distributed systems, neuronal networks, pattern recognition, soft-computing or data mining.