Romero Merino, Enrique
Total activity: 77
Research group
SOCO - Soft Computing
Department
Department of Software
School
Barcelona School of Informatics (FIB)
E-mail
eromerolsi.upc.edu
Contact details
UPC directory Open in new window

Graphic summary
  • Show / hide key
  • Information


Scientific and technological production
  •  

1 to 50 of 77 results
  • Access to the full text
    Identifying useful human feedback from an on-line translation service  Open access

     Barron Cedeño, Luis Alberto; Màrquez Villodre, Lluís; Henriquez Quintana, Carlos Alberto; Formiga Fanals, Lluis; Romero Merino, Enrique; May, Jonathan
    International Joint Conference on Artificial Intelligence
    Presentation's date: 2013-08-07
    Presentation of work at congresses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    Post-editing feedback provided by users of on-line translation services offers an excellent opportunity for automatic improvement of statistical machine translation (SMT) systems. However, feedback provided by casual users is very noisy, and must be automatically filtered in order to identify the poten- tially useful cases. We present a study on automatic feedback filtering in a real weblog collected from Reverso.net. We extend and re-annotate a train- ing corpus, define an extended set of simple fea- tures and approach the problem as a binary classi- fication task, experimenting with linear and kernel- based classifiers and feature selection. Results on the feedback filtering task show a significant im- provement over the majority class, but also a preci- sion ceiling around 70-80%. This reflects the inher- ent difficulty of the problem and indicates that shal- low features cannot fully capture the semantic na- ture of the problem. Despite the modest results on the filtering task, the classifiers are proven effective in an application-based evaluation. The incorpora- tion of a filtered set of feedback instances selected from a larger corpus significantly improves the per- formance of a phrase-based SMT system, accord- ing to a set of standard evaluation metrics

    Post-editing feedback provided by users of on-line translation services offers an excellent opportunity for automatic improvement of statistical machine translation (SMT) systems. However, feedback provided by casual users is very noisy, and must be automatically filtered in order to identify the potentially useful cases. We present a study on automatic feedback filtering in a real weblog collected from Reverso.net. We extend and re-annotate a training corpus, define an extended set of simple features and approach the problem as a binary classification task, experimenting with linear and kernelbased classifiers and feature selection. Results on the feedback filtering task show a significant improvement over the majority class, but also a precision ceiling around 70-80%. This reflects the inherent difficulty of the problemand indicates that shallow features cannot fully capture the semantic nature of the problem. Despite the modest results on the filtering task, the classifiers are proven effective in an application-based evaluation. The incorporation of a filtered set of feedback instances selected from a larger corpus significantly improves the performance of a phrase-based SMT system, according to a set of standard evaluation metrics.

  • A Quotient Basis Kernel for the prediction of mortality in severe sepsis patients

     Ribas Ripoll, Vicent; Romero Merino, Enrique; Ruiz Rodriguez, Juan Carlos; Vellido Alcacena, Alfredo
    European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
    Presentation's date: 2013-04-24
    Presentation of work at congresses

    Read the abstract Read the abstract View View Open in new window  Share Reference managers Reference managers Open in new window

    In this paper, we describe a novel kernel for multinomial distributions, namely the Quotient Basis Kernel (QBK), which is based on a suitable reparametrization of the input space through algebraic geometry and statistics. The QBK is used here for data transformation prior to classification in a medical problem concerning the prediction of mortality in patients suffering severe sepsis. This is a common clinical syndrome, often treated at the Intensive Care Unit (ICU) in a time-critical context. Mortality prediction results with Support Vector Machines using QBK compare favorably with those obtained using alternative kernels and standard clinical procedures.

  • A novel semi-supervised methodology for extracting tumor type-specific MRS sources in human brain data

     Ortega Martorell, Sandra; Riuz, Hector; Vellido Alcacena, Alfredo; Olier, Ivan; Romero Merino, Enrique; Julia Sape, Margarida; Martin, Jose D.; Harman, Ian H.; Arus, Carles; Lisboa, Paulo J G
    PLoS One
    Date of publication: 2013-12-23
    Journal article

    Read the abstract Read the abstract View View Open in new window  Share Reference managers Reference managers Open in new window

    Background: The clinical investigation of human brain tumors often starts with a non-invasive imaging study, providing information about the tumor extent and location, but little insight into the biochemistry of the analyzed tissue. Magnetic Resonance Spectroscopy can complement imaging by supplying a metabolic fingerprint of the tissue. This study analyzes single-voxel magnetic resonance spectra, which represent signal information in the frequency domain. Given that a single voxel may contain a heterogeneous mix of tissues, signal source identification is a relevant challenge for the problem of tumor type classification from the spectroscopic signal. Methodology/Principal Findings: Non-negative matrix factorization techniques have recently shown their potential for the identification of meaningful sources from brain tissue spectroscopy data. In this study, we use a convex variant of these methods that is capable of handling negatively-valued data and generating sources that can be interpreted as tumor class prototypes. A novel approach to convex non-negative matrix factorization is proposed, in which prior knowledge about class information is utilized in model optimization. Class-specific information is integrated into this semi-supervised process by setting the metric of a latent variable space where the matrix factorization is carried out. The reported experimental study comprises 196 cases from different tumor types drawn from two international, multi-center databases. The results indicate that the proposed approach outperforms a purely unsupervised process by achieving near perfect correlation of the extracted sources with the mean spectra of the tumor types. It also improves tissue type classification...

  • On the Intelligent Management of Sepsis in the Intensive Care Unit  Open access

     Ribas Ripoll, Vicente Jorge
    Defense's date: 2013-01-29
    Department of Software, Universitat Politècnica de Catalunya
    Theses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    The management of the Intensive Care Unit (ICU) in a hospital has its own, very specific requirements that involve, amongst others, issues of risk-adjusted mortality and average length of stay; nurse turnover and communication with physicians; technical quality of care; the ability to meet patient's family needs; and avoid medical error due rapidly changing circumstances and work overload. In the end, good ICU management should lead to an improvement in patient outcomes. Decision making at the ICU environment is a real-time challenge that works according to very tight guidelines, which relate to often complex and sensitive research ethics issues. Clinicians in this context must act upon as much available information as possible, and could therefore, in general, benefit from at least partially automated computer-based decision support based on qualitative and quantitative information. Those taking executive decisions at ICUs will require methods that are not only reliable, but also, and this is a key issue, readily interpretable. Otherwise, any decision tool, regardless its sophistication and accuracy, risks being rendered useless. This thesis addresses this through the design and development of computer based decision making tools to assist clinicians at the ICU. It focuses on one of the main problems that they must face: the management of the Sepsis pathology. Sepsis is one of the main causes of death for non-coronary ICU patients. Its mortality rate can reach almost up to one out of two patients for septic shock, its most acute manifestation. It is a transversal condition affecting people of all ages. Surprisingly, its definition has only been standardized two decades ago as a systemic inflammatory response syndrome with confirmed infection. The research reported in this document deals with the problem of Sepsis data analysis in general and, more specifically, with the problem of survival prediction for patients affected with Severe Sepsis. The tools at the core of the investigated data analysis procedures stem from the fields of multivariate and algebraic statistics, algebraic geometry, machine learning and computational intelligence. Beyond data analysis itself, the current thesis makes contributions from a clinical point of view, as it provides substantial evidence to the debate about the impact of the preadmission use of statin drugs in the ICU outcome. It also sheds light into the dependence between Septic Shock and Multi Organic Dysfunction Syndrome. Moreover, it defines a latent set of Sepsis descriptors to be used as prognostic factors for the prediction of mortality and achieves an improvement on predictive capability over indicators currently in use.

    La gestió d'una Unitat de Cures Intensives (UCI) hospitalària presenta uns requisits força específics incloent, entre altres, la disminució de la taxa de mortalitat, la durada de l'ingrès, la rotació d'infermeres i la comunicació entre metges amb al finalitad de donar una atenció de qualitat atenent als requisits tant dels malalts com dels familiars. També és força important controlar i minimitzar els error mèdics deguts a canvis sobtats i a la presa ràpida de deicisions assistencials. Al cap i a la fi, la bona gestió de la UCI hauria de resultar en una reducció de la mortalitat i durada d'estada. La presa de decisions en un entorn de crítics suposa un repte de presa de decisions en temps real d'acord a unes guies clíniques molt restrictives i que, pel que fa a la recerca, poden resultar en problemes ètics força sensibles i complexos. Per tant, el personal sanitari que ha de prendre decisions sobre la gestió de malalts crítics no només requereix eines de suport a la decisió que siguin fiables sinó que, a més a més, han de ser interpretables. Altrament qualsevol eina de decisió que no presenti aquests trets no és considerarà d'utilitat clínica. Aquesta tesi doctoral adreça aquests requisits mitjançant el desenvolupament d'eines de suport a la decisió per als intensivistes i es focalitza en un dels principals problemes als que s'han denfrontar: el maneig del malalt sèptic. La Sèpsia és una de les principals causes de mortalitats a les UCIS no-coronàries i la seva taxa de mortalitat pot arribar fins a la meitat dels malalts amb xoc sèptic, la seva manifestació més severa. La Sèpsia és un síndrome transversal, que afecta a persones de totes les edats. Sorprenentment, la seva definició ha estat estandaritzada, fa només vint anys, com a la resposta inflamatòria sistèmica a una infecció corfimada. La recerca presentada en aquest document fa referència a l'anàlisi de dades de la Sèpsia en general i, de forma més específica, al problema de la predicció de la supervivència de malalts afectats amb Sèpsia Greu. Les eines i mètodes que formen la clau de bòveda d'aquest treball provenen de diversos camps com l'estadística multivariant i algebràica, geometria algebraica, aprenentatge automàtic i inteligència computacional. Més enllà de l'anàlisi per-se, aquesta tesi també presenta una contribució des de el punt de vista clínic atès que presenta evidència substancial en el debat sobre l'impacte de l'administració d'estatines previ a l'ingrès a la UCI en els malalts sèptics. També s'aclareix la forta dependència entre el xoc sèptic i el Síndrome de Disfunció Multiorgànica. Finalment, també es defineix un conjunt de descriptors latents de la Sèpsia com a factors de pronòstic per a la predicció de la mortalitat, que millora sobre els mètodes actualment més utilitzats en la UCI.

  • Towards interpretable classifiers with blind signal separation

     Ruiz, Hector; Ortega Martorell, Sandra; Jarman, Ian H.; Vellido Alcacena, Alfredo; Martin Guerrero, Jose D.; Romero Merino, Enrique; Lisboa, Paulo J.G.
    International Conference on Neural Networks
    Presentation's date: 2012
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Cohort-based kernel visualisation with scatter matrices

     Romero Merino, Enrique; Mu, Tingting; Lisboa, Paulo J.G.
    Pattern recognition
    Date of publication: 2012-04
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Comparing error minimized extreme learning machines and support vector sequential feed-forward neural networks

     Romero Merino, Enrique; Alquezar Mancho, Renato
    Neural networks
    Date of publication: 2012-01
    Journal article

    Read the abstract Read the abstract View View Open in new window  Share Reference managers Reference managers Open in new window

    Recently, error minimized extreme learning machines (EM-ELMs) have been proposed as a simple and efficient approach to build single-hidden-layer feed-forward networks (SLFNs) sequentially. They add random hidden nodes one by one (or group by group) and update the output weights incrementally to minimize the sum-of-squares error in the training set. Other very similar methods that also construct SLFNs sequentially had been reported earlier with the main difference that their hidden-layer weights are a subset of the data instead of being random. These approaches are referred to as support vector sequential feed-forward neural networks (SV-SFNNs), and they are a particular case of the sequential approximation with optimal coefficients and interacting frequencies (SAOCIF) method. In this paper, it is firstly shown that EM-ELMs can also be cast as a particular case of SAOCIF. In particular, EM-ELMs can easily be extended to test some number of random candidates at each step and select the best of them, as SAOCIF does. Moreover, it is demonstrated that the cost of the computation of the optimal outputlayer weights in the originally proposed EM-ELMs can be improved if it is replaced by the one included in SAOCIF. Secondly, we present the results of an experimental study on 10 benchmark classification and 10 benchmark regression data sets, comparing EM-ELMs and SV-SFNNs, that was carried out under the same conditions for the two models. Although both models have the same (efficient) computational cost, a statistically significant improvement in generalization performance of SV-SFNNs vs. EM-ELMs was found in 12 out of the 20 benchmark problems.

  • Classification of human brain tumours from MRS data using Discrete Wavelet Transform and Bayesian Neural Networks

     Arizmendi Pereira, Carlos Julio; Vellido Alcacena, Alfredo; Romero Merino, Enrique
    Expert systems with applications
    Date of publication: 2012-04
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Data and knowledge visualization with virtual reality spaces, neural networks and rough sets: appplication to cancer and geophysical prospecting data

     Valdes Ramos, Julio Jose; Romero Merino, Enrique; Barton, Alan J.
    Expert systems with applications
    Date of publication: 2012-12-15
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Preprocessing MRS information for classification of human brain tumours

     Arizmendi Pereira, Carlos Julio; Vellido Alcacena, Alfredo; Romero Merino, Enrique
    Date of publication: 2012-06
    Book chapter

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • SIGNAL PROCESSING TECHNIQUES FOR BRAIN TUMOUR DIAGNOSIS FROM MAGNETIC RESONANCE SPECTROSCOPY DATA

     Arizmendi Pereira, Carlos Julio
    Defense's date: 2012-02-10
    Department of Software, Universitat Politècnica de Catalunya
    Theses

     Share Reference managers Reference managers Open in new window

  • Robust discrimination of glioblastomas from metastatic brain tumors on the basis of single-voxel 1H MRS

     Vellido Alcacena, Alfredo; Romero Merino, Enrique; Julià Sapé, Margarida; Majós, C.; Moreno Torres, À.; Pujol, Jesus; Arús, Carles
    NMR in biomedicine
    Date of publication: 2012-06
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Using the Leader algorithm with Support Vector Machines for large data sets

     Romero Merino, Enrique
    International Conference on Artificial Neural Networks
    Presentation's date: 2011-06
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Brain tumour classification using Gaussian decomposition and neural networks

     Arizmendi Pereira, Carlos Julio; Sierra, Daniel A.; Vellido Alcacena, Alfredo; Romero Merino, Enrique
    IEEE Engineering in Medicine and Biology Society
    Presentation's date: 2011-08-30
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Binary classification of brain tumours using a discrete wavelet transform and energy criteria

     Arizmendi Pereira, Carlos Julio; Vellido Alcacena, Alfredo; Romero Merino, Enrique
    Latin American Symposium on Circuits and Systems
    Presentation's date: 2011-02
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Cohort-based kernel visualisation with scatter matrices

     Romero Merino, Enrique; Fernandes, Ana Sofia; Mu, Tingting; Lisboa, Paulo J.G.
    International Conference on Neural Networks
    Presentation's date: 2010-07-21
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Diagnosis of brain tumours from magnetic resonance spectroscopy using wavelets and neural networks

     Arizmendi Pereira, Carlos Julio; Hernández-Tamames, Juan; Romero Merino, Enrique; Vellido Alcacena, Alfredo; Pozo, Francisco del
    IEEE Engineering in Medicine and Biology Society
    Presentation's date: 2010-09-02
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Feature and model selection with discriminatory visualization for diagnostic classification of brain tumors

     González Navarro, Félix Fernando; Belanche Muñoz, Luis Antonio; Romero Merino, Enrique; Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Arús, Carles
    Neurocomputing
    Date of publication: 2010-10
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Frequency selection for the diagnostic characterization of human brain tumours

     Arizmendi Pereira, Carlos Julio; Vellido Alcacena, Alfredo; Romero Merino, Enrique
    International Conference of the Catalan Association for Artificial Intelligence
    Presentation's date: 2009-10-22
    Presentation of work at congresses

     Share Reference managers Reference managers Open in new window

  • Data mining of patients on weaning trials from mechanical ventilation using cluster analysis and neural networks

     Arizmendi Pereira, Carlos Julio; Romero Merino, Enrique; Alquezar Mancho, Renato; Caminal Magrans, Pedro; Díaz, Ivan; Benito, Salvador; Giraldo Giraldo, Beatriz F.
    IEEE Engineering in Medicine and Biology Society
    Presentation's date: 2009-09
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Data mining of patients on weaning trials from mechanical ventilation using cluster analysis and neural networks

     Arizmendi, Carlos; Romero Merino, Enrique; Alquezar Mancho, Renato; Caminal Magrans, Pedro; Diaz, I; Benito ., Salvador; Giraldo Giraldo, Beatriz F.
    Annual International Conference of the IEEE Engineering in Medicine and Biology Society
    Date of publication: 2009
    Journal article

     Share Reference managers Reference managers Open in new window

  • Outlier exploration and diagnostic classification of a multi-centre 1H-MRS brain tumour database

     Vellido Alcacena, Alfredo; Romero Merino, Enrique; González Navarro, Félix Fernando; Belanche Muñoz, Luis Antonio; Julià Sapé, Margarida; Arús, Carles
    Neurocomputing
    Date of publication: 2009
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Exploratory characterization of a multi-centre 1H-MRS brain tumour database

     Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Romero Merino, Enrique; Arús, Carles
    Date of publication: 2009-01-31
    Book chapter

    Read the abstract Read the abstract View View Open in new window  Share Reference managers Reference managers Open in new window

    Machine learning is a powerful paradigm within which to analyze 1HMRS spectral data for the classification of tumour pathologies. An important characteristic of this task is the high dimensionality of the involved data sets. In this work we apply specific feature selection methods in order to reduce the complexity of the problem on two types of 1H-MRS spectral data: long-echo and short-echo time, which present considerable differences in the spectrum for the same cases. The experimental findings show that the feature selection methods enhance the classification performance of the models induced by several off-the-shelf classifiers and are able to offer very attractive solutions both in terms of prediction accuracy and number of involved spectral frequencies.

  • Generative manifold learning for the exploration of partially labeled data  Open access

     Cruz Barbosa, Raul
    Defense's date: 2009-10-01
    Department of Software, Universitat Politècnica de Catalunya
    Theses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    In many real-world application problems, the availability of data labels for supervised learning is rather limited. Incompletely labeled datasets are common in many of the databases generated in some of the currently most active areas of research. It is often the case that a limited number of labeled cases is accompanied by a larger number of unlabeled ones. This is the setting for semi-supervised learning, in which unsupervised approaches assist the supervised problem and vice versa. A manifold learning model, namely Generative Topographic Mapping (GTM), is the basis of the methods developed in this thesis. The non-linearity of the mapping that GTM generates makes it prone to trustworthiness and continuity errors that would reduce the faithfulness of the data representation, especially for datasets of convoluted geometry. In this thesis, a variant of GTM that uses a graph approximation to the geodesic metric is first defined. This model is capable of representing data of convoluted geometries. The standard GTM is here modified to prioritize neighbourhood relationships along the generated manifold. This is accomplished by penalizing the possible divergences between the Euclidean distances from the data points to the model prototypes and the corresponding geodesic distances along the manifold. The resulting Geodesic GTM (Geo-GTM) model is shown to improve the continuity and trustworthiness of the representation generated by the model, as well as to behave robustly in the presence of noise. The thesis then leads towards the definition and development of semi-supervised versions of GTM for partially-labeled data exploration. As a first step in this direction, a two-stage clustering procedure that uses class information is presented. A class information-enriched variant of GTM, namely class-GTM, yields a first cluster description of the data. The number of clusters defined by GTM is usually large for visualization purposes and does not necessarily correspond to the overall class structure. Consequently, in a second stage, clusters are agglomerated using the K-means algorithm with different novel initialization strategies that benefit from the probabilistic definition of GTM. We evaluate if the use of class information influences cluster-wise class separability. A robust variant of GTM that detects outliers while effectively minimizing their negative impact in the clustering process is also assessed in this context. We then proceed to the definition of a novel semi-supervised model, SS-Geo-GTM, that extends Geo-GTM to deal with semi-supervised problems. In SS-Geo-GTM, the model prototypes are linked by the nearest neighbour to the data manifold constructed by Geo-GTM. The resulting proximity graph is used as the basis for a class label propagation algorithm. The performance of SS-Geo-GTM is experimentally assessed, comparing positively with that of an Euclidean distance-based counterpart and that of the alternative Laplacian Eigenmaps method. Finally, the developed models (the two-stage clustering procedure and the semi-supervised models) are applied to the analysis of a human brain tumour dataset (obtained by Nuclear Magnetic Resonance Spectroscopy), where the tasks are, in turn, data clustering and survival prognostic modeling.

    Resum de la tesi (màxim 4000 caràcters. Si se supera aquest límit, el resum es tallarà automàticament al caràcter 4000) En muchos problemas de aplicación del mundo real, la disponibilidad de etiquetas de datos para aprendizaje supervisado es bastante limitada. La existencia de conjuntos de datos etiquetados de manera incompleta es común en muchas de las bases de datos generadas en algunas de las áreas de investigación actualmente más activas. Es frecuente que un número limitado de casos etiquetados venga acompañado de un número mucho mayor de datos no etiquetados. Éste es el contexto en el que opera el aprendizaje semi-supervisado, en el cual enfoques no-supervisados prestan ayuda a problemas supervisados y vice versa. Un modelo de aprendizaje de variaciones (manifold learning, en inglés), llamado Mapeo Topográfico Generativo (GTM, en acrónimo de su nombre en inglés), es la base de los métodos desarrollados en esta tesis. La no-linealidad del mapeo que GTM genera hace que éste sea propenso a errores de fiabilidad y continuidad, los cuales pueden reducir la fidelidad de la representación de los datos, especialmente para conjuntos de datos de geometría intrincada. En esta tesis, una extensión de GTM que utiliza una aproximación vía grafos a la métrica geodésica es definida en primer lugar. Este modelo es capaz de representar datos con geometrías intrincadas. En él, el GTM estándar es modificado para priorizar relaciones de vecindad a lo largo de la variación generada. Esto se logra penalizando las divergencias existentes entre las distancias Euclideanas de los datos a los prototipos del modelo y las correspondientes distancias geodésicas a lo largo de la variación. Se muestra que el modelo Geo-GTM resultante mejora la continuidad y fiabilidad de la representación generada y que se comporta de manera robusta en presencia de ruido. Más adelante, la tesis nos lleva a la definición y desarrollo de versiones semi-supervisadas de GTM para la exploración de conjuntos de datos parcialmente etiquetados. Como un primer paso en esta dirección, se presenta un procedimiento de agrupamiento en dos etapas que utiliza información de pertenencia a clase. Una extensión de GTM enriquecida con información de pertenencia a clase, llamada class-GTM, produce una primera descripción de grupos de los datos. El número de grupos definidos por GTM es normalmente grande para propósitos de visualización y no corresponde necesariamente con la estructura de clases global. Por ello, en una segunda etapa, los grupos son aglomerados usando el algoritmo K-means con diferentes estrategias de inicialización novedosas las cuales se benefician de la definición probabilística de GTM. Evaluamos si el uso de información de clase influye en la separabilidad de clase por grupos. Una extensión robusta de GTM que detecta datos atípicos a un tiempo que minimiza de forma efectiva su impacto negativo en el proceso de agrupamiento es evaluada también en este contexto. Se procede después a la definición de un nuevo modelo semi-supervisado, SS-Geo-GTM, que extiende Geo-GTM para ocuparse de problemas semi-supervisados. En SS-Geo-GTM, los prototipos del modelo son vinculados al vecino más cercano a la variación construída por Geo-GTM. El grafo de proximidad resultante es utilizado como base para un algoritmo de propagación de etiquetas de clase. El rendimiento de SS-Geo-GTM es valorado experimentalmente, comparando positivamente tanto con la contraparte de este modelo basada en la distancia Euclideana como con el método alternativo Laplacian Eigenmaps. Finalmente, los modelos desarrollados (el procedimiento de agrupamiento en dos etapas y los modelos semi-supervisados) son aplicados al análisis de un conjunto de datos de tumores cerebrales humanos (obtenidos mediante Espectroscopia de Resonancia Magnética Nuclear), donde las tareas a realizar son el agrupamiento de datos y el modelado de pronóstico de supervivencia.

  • An experimental study on methods for the selection of basis functions in regression

     Barrio Moliner, Ignacio; Romero Merino, Enrique; Belanche Muñoz, Luis Antonio
    Neurocomputing
    Date of publication: 2009-08
    Journal article

     Share Reference managers Reference managers Open in new window

  • Feature Selection with Single-Layer Perceptrons for a multicentre 1H-MRS brain tumour database

     Romero Merino, Enrique; Vellido Alcacena, Alfredo; Sopena, Josep Maria
    Lecture notes in computer science
    Date of publication: 2009-06-12
    Journal article

    Read the abstract Read the abstract View View Open in new window  Share Reference managers Reference managers Open in new window

    A Feature Selection process with Single-Layer Perceptrons is shown to provide optimum discrimination of an international, multi-centre 1H-MRS database of brain tumors at reasonable computational cost. Results are both intuitively interpretable and very accurate. The method remains simple enough as to allow its easy integration in existing medical decision support systems.

  • Patients on weaning trials classified with neural networks and feature selection

     Giraldo Giraldo, Beatriz F.; Arizmendi, C; Romero Merino, Enrique; Alquezar Mancho, Renato; Caminal Magrans, Pedro; Benito ., Salvador
    Date of publication: 2008-09
    Book chapter

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Neural Network Based Visual Data Mining for Cancer Data

     Romero Merino, Enrique
    Date of publication: 2008-08-31
    Book chapter

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • DSS-oriented Exploration of a Multi-centre Magnetic Resonance Spectroscopy Brain Tumour Dataset Through Visualization

     Romero Merino, Enrique; Julià-Sapé, M; Vellido Alcacena, Alfredo
    European Symposium on Artificial Neural Networks
    Presentation of work at congresses

     Share Reference managers Reference managers Open in new window

  • Access to the full text
    Classification, dimensionality reduction, and maximally discriminatory visualization of a multicentre 1H-MRS database of brain tumors  Open access

     Lisboa, Paulo J.G.; Romero Merino, Enrique; Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Arús, Carles
    IEEE International Conference on Machine Learning and Applications
    Presentation's date: 2008
    Presentation of work at congresses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    The combination of an Artificial Neural Network classifier, a feature selection process, and a novel linear dimensionality reduction technique that provides a data projection for visualization and which preserves completely the class discrimination achieved by the classifier, is applied in this study to the analysis of an international, multi-centre 1H-MRS database of brain tumors. This combination yields results that are both intuitively interpretable and very accurate. The method as a whole remains simple enough as to allow its easy integration in existing medical decision support systems.

  • Performing Feature Selection with Multi-Layer Perceptrons

     Romero Merino, Enrique
    IEEE transactions on neural networks
    Date of publication: 2008-03
    Journal article

     Share Reference managers Reference managers Open in new window

  • Exploratory characterization of outliers in a multi-centre 1H-MRS brain tumour dataset

     Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Romero Merino, Enrique; Arús, Carles
    Lecture notes in computer science
    Date of publication: 2008-09
    Journal article

    Read the abstract Read the abstract View View Open in new window  Share Reference managers Reference managers Open in new window

    As part of the AIDTumour research project, the analysis of MRS data corresponding to various tumour pathologies is used to assist expert diagnosis. The high dimensionality of the MR spectra might obscure atypical aspects of the data that would jeopardize their automated classification and, as a result, the process of computer-based diagnostic assistance. In this paper, we put forward a method to overcome this potential problem that combines automatic outlier detection, visualization through dimensionality reduction, and expert opinion.

  • Data and Knowledge Visualization with Virtual Reality Spaces, Neural Networks and Rough Sets: Application to Geophysical Prospecting

     Valdés, J J; Romero Merino, Enrique; González, R
    International Joint Conference on Neural Networks
    Presentation of work at congresses

     Share Reference managers Reference managers Open in new window

  • Neural Network Based Virtual Reality Spaces for Visual Data Mining of Cancer Data: An Unsupervised Perspective

     Romero Merino, Enrique
    Lecture notes in computer science
    Date of publication: 2007-06
    Journal article

     Share Reference managers Reference managers Open in new window

  • Extended linear models with gaussian prior on the parameters and adaptive expansion vectors

     Barrio Moliner, Ignacio; Romero Merino, Enrique; Belanche Muñoz, Luis Antonio
    Lecture notes in computer science
    Date of publication: 2007-09
    Journal article

     Share Reference managers Reference managers Open in new window

  • Selection of basis functions guided by the L2 soft margin

     Barrio Moliner, Ignacio; Romero Merino, Enrique; Belanche Muñoz, Luis Antonio
    Lecture notes in computer science
    Date of publication: 2007-09
    Journal article

     Share Reference managers Reference managers Open in new window

  • Comparing support vector machines and feedforward neural networks with similar hidden-layer weights

     Romero Merino, Enrique; Toppo, D
    IEEE transactions on neural networks
    Date of publication: 2007-05
    Journal article

     Share Reference managers Reference managers Open in new window

  • Incremental and decremental learning for linear support vector machines

     Romero Merino, Enrique; Barrio Moliner, Ignacio; Belanche Muñoz, Luis Antonio
    Lecture notes in computer science
    Date of publication: 2007-09
    Journal article

     Share Reference managers Reference managers Open in new window

  • Heuristics for the selection of weights in sequential feed-forward neural networks: An experimental study

     Romero Merino, Enrique; Alquezar Mancho, Renato
    Neurocomputing
    Date of publication: 2007-10
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Patients on weaning trials from mechanical ventilation classified with neural networks and feature selection

     Giraldo Giraldo, Beatriz F.; Arizmendi, Carlos; Romero Merino, Enrique; Alquezar Mancho, Renato; Caminal Magrans, Pedro; Benito ., Salvador; Ballesteros Carrillo, David
    IEEE Engineering in Medicine and Biology Society
    Presentation's date: 2006-08-30
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Clasificación de pacientes en proceso de extubacion mediante redes neuronales y selección características

     Giraldo Giraldo, Beatriz F.; Arizmendi, C; Romero Merino, Enrique; Alquezar Mancho, Renato; Caminal Magrans, Pedro; Benito ., Salvador; Ballesteros Carrillo, David
    Congreso Anual de la Sociedad Española de Ingeniería Biomédica
    Presentation's date: 2006-11-06
    Presentation of work at congresses

     Share Reference managers Reference managers Open in new window

  • Comparing Support Vector Machines and Feed-forward Neural Networks with Similar Parameters

     Romero Merino, Enrique
    Lecture notes in computer science
    Date of publication: 2006-09
    Journal article

     Share Reference managers Reference managers Open in new window

  • A sequential algorithm for feed-forward neural networks with optimal coefficients and interacting frequencies

     Romero Merino, Enrique; Alquezar Mancho, Renato
    Neurocomputing
    Date of publication: 2006-08
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Feature selection and outliers detection with genetic algorithms and neural networks

     Solanas, Agusti; Romero Merino, Enrique; Gómez, Sergio; Alquezar Mancho, Renato; Domingo Ferrer, Josep; Sopena, Josep Maria
    Eighth Catalan Conference on Artificial Intelligence
    Presentation's date: 2005-10-26
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Microestructural changes of compacted bentonita induced by hydro-mechanical actions

     Suriol Castellvi, Jose; Lloret Morancho, Antonio; Romero Merino, Enrique; Hoffmann, C; Castellanos, E
    Date of publication: 2005-01
    Book chapter

     Share Reference managers Reference managers Open in new window

  • Feature selection and outliers detection with genetic algorithms and neural networks

     Solanas, Agusti; Romero Merino, Enrique; Gómez, Sergio; Domingo Ferrer, Josep; Alquezar Mancho, Renato; Sopena, Josep Maria
    Date of publication: 2005-10
    Book chapter

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • An Experimental Study of Several Decision Issues for Feature Selection with Multi-Layer Perceptrons

     Romero Merino, Enrique; Sopena, J M
    International Joint Conference on Neural Networks
    Presentation of work at congresses

     Share Reference managers Reference managers Open in new window

  • An Experimental Study of Several Decision Issues for Feature Selection with Multi-Layer Perceptrons

     Romero Merino, Enrique
    International Joint Conference on Neural Networks
    Presentation's date: 2005-08-01
    Presentation of work at congresses

     Share Reference managers Reference managers Open in new window

  • Selection of Weights for Sequential Feed-forward Neural Networks: An Experimental Study

     Romero Merino, Enrique
    Lecture notes in computer science
    Date of publication: 2005-06
    Journal article

     Share Reference managers Reference managers Open in new window

  • Learning with feed-forward neural networks: Three schemes to deal with the bias/variance trade-off  Open access

     Romero Merino, Enrique
    Defense's date: 2004-11-30
    Department of Software, Universitat Politècnica de Catalunya
    Theses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    In terms of the Bias/Variance decomposition, very flexible (i.e., complex) Supervised Machine Learning systems may lead to unbiased estimators but with high variance. A rigid model, in contrast, may lead to small variance but high bias. There is a trade-off between the bias and variance contributions to the error, where the optimal performance is achieved.In this work we present three schemes related to the control of the Bias/Variance decomposition for Feed-forward Neural Networks (FNNs) with the (sometimes modified) quadratic loss function:1. An algorithm for sequential approximation with FNNs, named Sequential Approximation with Optimal Coefficients and Interacting Frequencies (SAOCIF). Most of the sequential approximations proposed in the literature select the new frequencies (the non-linear weights) guided by the approximation of the residue of the partial approximation. We propose a sequential algorithm where the new frequency is selected taking into account its interactions with the previously selected ones. The interactions are discovered by means of their optimal coefficients (the linear weights). A number of heuristics can be used to select the new frequencies. The aim is that the same level of approximation may be achieved with less hidden units than if we only try to match the residue as best as possible. In terms of the Bias/Variance decomposition, it will be possible to obtain simpler models with the same bias. The idea behind SAOCIF can be extended to approximation in Hilbert spaces, maintaining orthogonal-like properties. In this case, the importance of the interacting frequencies lies in the expectation of increasing the rate of approximation. Experimental results show that the idea of interacting frequencies allows to construct better approximations than matching the residue.2. A study and comparison of different criteria to perform Feature Selection (FS) with Multi-Layer Perceptrons (MLPs) and the Sequential Backward Selection (SBS) procedure within the wrapper approach. FS procedures control the Bias/Variance decomposition by means of the input dimension, establishing a clear connection with the curse of dimensionality. Several critical decision points are studied and compared. First, the stopping criterion. Second, the data set where the value of the loss function is measured. Finally, we also compare two ways of computing the saliency (i.e., the relative importance) of a feature: either first train a network and then remove temporarily every feature or train a different network with every feature temporarily removed. The experiments are performed for linear and non-linear models. Experimental results suggest that the increase in the computational cost associated with retraining a different network with every feature temporarily removed previous to computing the saliency can be rewarded with a significant performance improvement, specially if non-linear models are used. Although this idea could be thought as very intuitive, it has been hardly used in practice. Regarding the data set where the value of the loss function is measured, it seems clear that the SBS procedure for MLPs takes profit from measuring the loss function in a validation set. A somewhat non-intuitive conclusion is drawn looking at the stopping criterion, where it can be seen that forcing overtraining may be as useful as early stopping.3. A modification of the quadratic loss function for classification problems, inspired in Support Vector Machines (SVMs) and the AdaBoost algorithm, named Weighted Quadratic Loss (WQL) function. The modification consists in weighting the contribution of every example to the total error. In the linearly separable case, the solution of the hard margin SVM also minimizes the proposed loss function. The hardness of the resulting solution can be controlled, as in SVMs, so that this scheme may also be used for the non-linearly separable case. The error weighting proposed in WQL forces the training procedure to pay more attention to the points with a smaller margin. Therefore, variance tries to be controlled by not attempting to overfit the points that are already well classified. The model shares several properties with the SVMs framework, with some additional advantages. On the one hand, the final solution is neither restricted to have an architecture with so many hidden units as points (or support vectors) in the data set nor to use kernel functions. The frequencies are not restricted to be a subset of the data set. On the other hand, it allows to deal with multiclass and multilabel problems in a natural way. Experimental results are shown confirming these claims.A wide experimental work has been done with the proposed schemes, including artificial data sets, well-known benchmark data sets and two real-world problems from the Natural Language Processing domain. In addition to widely used activation functions, such as the hyperbolic tangent or the Gaussian function, other activation functions have been tested. In particular, sinusoidal MLPs showed a very good behavior. The experimental results can be considered as very satisfactory. The schemes presented in this work have been found to be very competitive when compared to other existing schemes described in the literature. In addition, they can be combined among them, since they deal with complementary aspects of the whole learning process.