An Interval-based Multiobjective Approach to Feature Subset Selection Using Joint Modeling of Objectives and Variables

This paper studies feature subset selection in classification using a multiobjective estimation of distribution algorithm. We consider six functions, namely area under ROC curve, sensitivity, specificity, precision, F1 measure and Brier score, for evaluation of feature subsets and as the objectives of the problem. One of the characteristics of these objective functions is the existence of noise in their values that should be appropriately handled during optimization. Our proposed algorithm consists of two major techniques which are specially designed for the feature subset selection problem. The first one is a solution ranking method based on interval values to handle the noise in the objectives of this problem. The second one is a model estimation method for learning a joint probabilistic model of objectives and variables which is used to generate new solutions and advance through the search space. To simplify model estimation, l1 regularized regression is used to select a subset of problem variables before model learning. The proposed algorithm is compared with a well-known ranking method for interval-valued objectives and a standard multiobjective genetic algorithm. Particularly, the effects of the two new techniques are experimentally investigated. The experimental results show that the proposed algorithm is able to obtain comparable or better performance on the tested datasets.

Multi-objective Estimation of Distribution Algorithm Based on Joint Modeling of Objectives and Variables

This paper proposes a new multi-objective estimation of distribution algorithm (EDA) based on joint modeling of objectives and variables. This EDA uses the multi-dimensional Bayesian network as its probabilistic model. In this way it can capture the dependencies between objectives, variables and objectives, as well as the dependencies learnt between variables in other Bayesian network-based EDAs. This model leads to a problem decomposition that helps the proposed algorithm to find better trade-off solutions to the multi-objective problem. In addition to Pareto set approximation, the algorithm is also able to estimate the structure of the multi-objective problem. To apply the algorithm to many-objective problems, the algorithm includes four different ranking methods proposed in the literature for this purpose. The algorithm is applied to the set of walking fish group (WFG) problems, and its optimization performance is compared with an evolutionary algorithm and another multi-objective EDA. The experimental results show that the proposed algorithm performs significantly better on many of the problems and for different objective space dimensions, and achieves comparable results on some compared with the other algorithms.

Conformance monitoring, contract of objectives, tube in the sky

Ponencia invitada sobre asignacion y gestion de losts en el curso de verano de la UPM Research in Decisión Support Systems for future Air Traffic Management

Vacíos urbanos en la ciudad de Zaragoza (1975-2010). Oportunidades para la estructuración y continuidad urbana

La expansión de las ciudades hacia la periferia en las últimas décadas ha generado una serie de tejidos en colisión, donde los intersticios y los espacios residuales muestran la realidad de una ciudad dispersa. A esta serie de espacios intermedios los denominaremos “vacíos urbanos” y determinaremos una serie de factores que hagan posible su identificación Los vacíos urbanos son el objeto de la Tesis. El concepto “vacío urbano” adquiere un determinado significado para la investigación y se acota formulando una serie de parámetros para su definición. Definiremos los “vacíos urbanos” como espacios que han aparecido en el extrarradio de las ciudades, fruto de una expansión sin precedentes de las áreas urbanas. Han surgido como lugares residuales condicionados por elementos naturales o por infraestructuras, ligados a una temporalidad incierta, no inmersos en la dinámicas urbanas o habiendo perdido su funcionalidad. El objetivo de esta investigación es localizar y clasificar los vacíos existentes en la ciudad partiendo de la hipótesis de que es posible establecer una metodología para su reconocimiento y comprobación de los parámetros que los definen. La investigación centra su estudio en la ciudad de Zaragoza, como un ejemplo paradigmático, confirmando que la nueva fenomenología territorial, no sólo se manifiesta en dicha ciudad y en las áreas en concreto donde se ha estudiado, sino que más bien se trata de observarla como una serie de modelos tipológicos que respondan a un proceso de análisis. La propuesta metodológica de loa tesis pasa por reconocerlos, mostrarlos y darles una visibilidad que permita su clasificación, el desarrollo de un estudio de Áreas de la ciudad donde se localizan y un análisis de cada tipología de vacío urbano. Como instrumento metodológico se ha elaborado una detallada cartografía a diferentes escalas. La realización de los planos ha sido el medio de análisis que ha permitido localizar e interpretar los vacíos, complementado con unas escogidas fotografías de estos entornos. La metodología se convierte en un modo de descubrir y analizar los vacíos, el proceso conlleva la comprobación y la clasificación en una tipología. Según se han ido analizando nuevas áreas, se confirma como un hecho repetitivo exitoso y se observa que es aplicable a otras áreas urbanas de similares características, donde se den las condiciones principales de cada tipo. En general, deberán ser espacios localizados en la periferia de núcleos urbanos donde aparezcan fenómenos ligados a una expansión discontinua. Parece probable que el futuro de la ciudad se tenga que resolver en las próximas décadas sobre su actual extensión; bajo este supuesto, es trascendente la función que los vacíos puedan desempeñar en el futuro desarrollo urbano. Toda posibilidad de intervención precisará de una necesaria reinterpretación, puestas las miras en su potencial como elementos capaces de generar una rehabilitación urbana. A su vez se hace conveniente plantear una reflexión sobre estos espacios cargados de una dimensión social y cultural, como lugares capaces de articular y dotar de identidad al medio urbano. Los vacíos deberían protagonizar un papel relevante en la estructuración urbana, abriendo posibilidades para el tramado de la ciudad desde nuevas perspectivas. Una apuesta para que estos espacios libres pasen de ser el objetivo de procesos urbanizadores tradicionales, a ser considerados como oportunos elementos vertebradores de los entornos periurbanos, colaborando en el objetivo de una ciudad contemporánea sostenible. ABSTRACT In the last decades the city periphary expansion has provoked a series of matters to collide, where interstices and waste lands show the reality of a divided city. We shall determine a number of factors allowing us to treat these “in between” spaces, also called "urban voids" as identifiable elements. We will consider them the subject of the thesis and establish them as a "concept", delimiting the meaning of the research specifically acquired, defining, and formulating a set of parameters. "Urban voids" are defined as spaces that have appeared on the outskirts of cities, the result of an unprecedented expansion of urban areas. They have emerged as waste lands, conditioned by natural elements or infrastructure, related to uncertain temporality, not immersed in the urban dynamics, or having lost their functionality. It seems likely that in the upcoming decades, the future of the city will have to resolve its current way of expanding . It is under this assumption that urban voids, as intermediate spaces, will play an important role in future urban developments. Any possible intervention will require a necessary reinterpretation, closely watching their potential as elements capable of generating an urban rehabilitation. At the same time, we wish to reflect on these spaces, in many cases loaded with a social and cultural dimension, as places able to articulate and give identity to the urban environment. Based on the hypothesis that it is possible to establish a methodology for recognition and verification, the purpose of this research is to locate and classify the existing urban voids in the city. The research focuses its study on the city of Saragossa which can be seen as a paradigmatic example. The objective is to confirm that this new territorial phenomenology, not only manifests itself in Saragossa and specific areas under study, but also, can be observed as a series of typological models that respond to a review process. The methodological proposal will recognize, demonstrate, and exhibit these urban voids, in a light that will allow us to classify them, examine the different areas where they can be found, and develop a tipology analysis of each type of urban void found. The technical tool used in this research is a detailed mapping at different scales. A realization of plans, as a mean of analysis, supplemented with a few selected pictures of these environments, will facilitate the location and interpretation of these voids. This scholarly approach to discover and analyze urban voids will involve checking and classifying them in a typology. It has been confirmed as a successful constant regulator while exploring new areas. It will be apply to other urban areas with similar characteristics, that is, spaces located on the periphery of urban areas where expansion is linked to the appearence of discontinuous phenomena. The urban voids play a fundamental role within the urban structure, providing the city a weaving scheme with opportunities of fresh perspectives. The challenge for these free spaces is to move from being the consequence of traditional urban development processes, to being considered opportune backbone elements of peri-urban environments, and to finally contribute to the objectives of a sustainable contemporary city.