http://hdl.handle.net/2117/3687 Fri, 24 May 2013 00:12:32 GMT 2013-05-24T00:12:32Z webmaster.bupc@upc.edu Universitat Politècnica de Catalunya. Servei de Biblioteques i Documentació no http://hdl.handle.net/2117/17240 Title: Convex non-negative matrix factorization for brain tumor delimitation from MRSI data Authors: Ortega Martorell, Sandra; Lisboa, Paulo J.G.; Vellido Alcacena, Alfredo; Simoes, Rui V.; Pumarola, Martí; Julià Sapé, Margarida; Arús, Carles Wed, 09 Jan 2013 16:46:10 GMT http://hdl.handle.net/2117/17240 2013-01-09T16:46:10Z Ortega Martorell, Sandra; Lisboa, Paulo J.G.; Vellido Alcacena, Alfredo; Simoes, Rui V.; Pumarola, Martí; Julià Sapé, Margarida; Arús, Carles no http://hdl.handle.net/2117/16937 Title: A new graph matching method for point-set correspondence using the EM algorithm and Softassign Authors: Sanromà Güell, Gerard; Alquézar Mancho, René; Serratosa Casanelles, Francesc Abstract: Finding correspondences between two point-sets is a common step in many vision applications (e.g., image matching or shape retrieval). We present a graph matching method to solve the point-set correspondence problem, which is posed as one of mixture modelling. Our mixture model encompasses a model of structural coherence and a model of affine-invariant geometrical errors. Instead of absolute positions, the geometrical positions are represented as relative positions of the points with respect to each other. We derive the Expectation–Maximization algorithm for our mixture model. In this way, the graph matching problem is approximated, in a principled way, as a succession of assignment problems which are solved using Softassign. Unlike other approaches, we use a true continuous underlying correspondence variable. We develop effective mechanisms to detect outliers. This is a useful technique for improving results in the presence of clutter. We evaluate the ability of our method to locate proper matches as well as to recognize object categories in a series of registration and recognition experiments. Our method compares favourably to other graph matching methods as well as to point-set registration methods and outlier rejectors. Fri, 16 Nov 2012 11:41:25 GMT http://hdl.handle.net/2117/16937 2012-11-16T11:41:25Z Sanromà Güell, Gerard; Alquézar Mancho, René; Serratosa Casanelles, Francesc no Correspondence problem, Graph matching, Affine registration, Outlier detection, Expectation maximization, Softassign Finding correspondences between two point-sets is a common step in many vision applications (e.g., image matching or shape retrieval). We present a graph matching method to solve the point-set correspondence problem, which is posed as one of mixture modelling. Our mixture model encompasses a model of structural coherence and a model of affine-invariant geometrical errors. Instead of absolute positions, the geometrical positions are represented as relative positions of the points with respect to each other. We derive the Expectation–Maximization algorithm for our mixture model. In this way, the graph matching problem is approximated, in a principled way, as a succession of assignment problems which are solved using Softassign. Unlike other approaches, we use a true continuous underlying correspondence variable. We develop effective mechanisms to detect outliers. This is a useful technique for improving results in the presence of clutter. We evaluate the ability of our method to locate proper matches as well as to recognize object categories in a series of registration and recognition experiments. Our method compares favourably to other graph matching methods as well as to point-set registration methods and outlier rejectors. http://hdl.handle.net/2117/16872 Title: Comparing error minimized extreme learning machines and support vector sequential feed-forward neural networks Authors: Romero Merino, Enrique; Alquézar Mancho, René Abstract: 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. Fri, 09 Nov 2012 10:32:51 GMT http://hdl.handle.net/2117/16872 2012-11-09T10:32:51Z Romero Merino, Enrique; Alquézar Mancho, René no Error minimized extreme learning machines, Sequential approximations, Support vector sequential feed-forward neural networks 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. http://hdl.handle.net/2117/13279 Title: Variational Bayesian generative topographic mapping Authors: Olier Caparroso, Iván; Vellido Alcacena, Alfredo Abstract: General finite mixture models are powerful tools for the density-based grouping of multivariate i.i.d. data, but they lack data visualization capabilities, which reduces their practical applicability to real-world problems. Generative topographic mapping (GTM) was originally formulated as a constrained mixture of distributions in order to provide simultaneous visualization and clustering of multivariate data. In its inception, the adaptive parameters were determined by maximum likelihood (ML), using the expectation-maximization (EM) algorithm. The original GTM is, therefore, prone to data overfitting unless a regularization mechanism is included. In this paper, we define an alternative variational formulation of GTM that provides a full Bayesian treatment to a Gaussian process (GP)-based variation of the model. The generalization capabilities of the proposed Variational Bayesian GTM are assessed in some detail and compared with those of alternative GTM regularization approaches in terms of test log-likelihood, using several artificial and real datasets. Wed, 21 Sep 2011 10:24:23 GMT http://hdl.handle.net/2117/13279 2011-09-21T10:24:23Z Olier Caparroso, Iván; Vellido Alcacena, Alfredo no Variational methods, Generative topographic mapping, Clustering, Data visualization, Adaptive regularization, Overfitting General finite mixture models are powerful tools for the density-based grouping of multivariate i.i.d. data, but they lack data visualization capabilities, which reduces their practical applicability to real-world problems. Generative topographic mapping (GTM) was originally formulated as a constrained mixture of distributions in order to provide simultaneous visualization and clustering of multivariate data. In its inception, the adaptive parameters were determined by maximum likelihood (ML), using the expectation-maximization (EM) algorithm. The original GTM is, therefore, prone to data overfitting unless a regularization mechanism is included. In this paper, we define an alternative variational formulation of GTM that provides a full Bayesian treatment to a Gaussian process (GP)-based variation of the model. The generalization capabilities of the proposed Variational Bayesian GTM are assessed in some detail and compared with those of alternative GTM regularization approaches in terms of test log-likelihood, using several artificial and real datasets. http://hdl.handle.net/2117/13075 Title: New multiplatform computer program for numerical identification of microorganisms Authors: Flores, Oscar; Belanche Muñoz, Luis Antonio; Blanch, Anicet R. Abstract: The classification of bacteria by using genomic methods or expensive biochemical-based commercial kits is sometimes beyond the reach of many laboratories that need to perform numerous classifications of unknown bacterial strains in a fast, cheap, and reliable way. A new computer program, Identax, for the computer-assisted identification of microorganisms by using only results obtained from conventional biochemical tests is presented. Identax improves current microbial identification software and provides a multiplatform and userfriendly program. It can be executed from any operating system and can be downloaded without any cost from the Identax website (www.identax.org). Fri, 29 Jul 2011 08:44:34 GMT http://hdl.handle.net/2117/13075 2011-07-29T08:44:34Z Flores, Oscar; Belanche Muñoz, Luis Antonio; Blanch, Anicet R. no The classification of bacteria by using genomic methods or expensive biochemical-based commercial kits is sometimes beyond the reach of many laboratories that need to perform numerous classifications of unknown bacterial strains in a fast, cheap, and reliable way. A new computer program, Identax, for the computer-assisted identification of microorganisms by using only results obtained from conventional biochemical tests is presented. Identax improves current microbial identification software and provides a multiplatform and userfriendly program. It can be executed from any operating system and can be downloaded without any cost from the Identax website (www.identax.org). http://hdl.handle.net/2117/13071 Title: On the improvement of the mapping trustworthiness and continuity of a manifold learning model Authors: Cruz Barbosa, Raúl; Vellido Alcacena, Alfredo Abstract: Manifold learningmethodsmodel high-dimensional data through low-dimensional manifolds embedded in the observed data space. This simplification implies that their are prone to trustworthiness and continuity errors. Generative Topographic Mapping (GTM) is one such manifold learning method for multivariate data clustering and visualization, defined within a probabilistic framework. In the original formulation,GTMis optimized byminimization of an error that is a function of Euclidean distances, making it vulnerable to the aforementioned errors, especially for datasets of convoluted geometry. Here, we modify GTM to penalize divergences between theEuclidean distances fromthe datapoints to themodel prototypes and the corresponding geodesic distances along the manifold. Several experiments with artificial data showthat this strategy improves the continuity and trustworthiness of the data representation generated by the model. Thu, 28 Jul 2011 09:51:31 GMT http://hdl.handle.net/2117/13071 2011-07-28T09:51:31Z Cruz Barbosa, Raúl; Vellido Alcacena, Alfredo no Manifold learningmethodsmodel high-dimensional data through low-dimensional manifolds embedded in the observed data space. This simplification implies that their are prone to trustworthiness and continuity errors. Generative Topographic Mapping (GTM) is one such manifold learning method for multivariate data clustering and visualization, defined within a probabilistic framework. In the original formulation,GTMis optimized byminimization of an error that is a function of Euclidean distances, making it vulnerable to the aforementioned errors, especially for datasets of convoluted geometry. Here, we modify GTM to penalize divergences between theEuclidean distances fromthe datapoints to themodel prototypes and the corresponding geodesic distances along the manifold. Several experiments with artificial data showthat this strategy improves the continuity and trustworthiness of the data representation generated by the model. http://hdl.handle.net/2117/12331 Title: Advances in machine learning and computational intelligence Authors: Schleif, Frank-Michael; Biehl, Michael; Vellido Alcacena, Alfredo Mon, 11 Apr 2011 10:29:16 GMT http://hdl.handle.net/2117/12331 2011-04-11T10:29:16Z Schleif, Frank-Michael; Biehl, Michael; Vellido Alcacena, Alfredo no http://hdl.handle.net/2117/12330 Title: Advances in clustering and visualization of time series using GTM through time Authors: Olier Caparroso, Iván; Vellido Alcacena, Alfredo Abstract: Most of the existing research on multivariate time series concerns supervised forecasting problems. In comparison, little research has been devoted to their exploration through unsupervised clustering and visualization. In this paper, the capabilities of Generative Topographic Mapping Through Time, a model with foundations in probability theory, that performs simultaneous time series clustering and visualization, are assessed in detail. Focus is placed on the visualization of the evolution of signal regimes and the exploration of sudden transitions, for which a novel identification index is defined. The interpretability of time series clustering results may become extremely difficult, even in exploratory visualization, for high dimensional datasets. Here, we define and test an unsupervised time series relevance determination method, fully integrated in the Generative Topographic Mapping Through Time model, that can be used as a basis for time series selection. This method should ease the interpretation of time series clustering results. Mon, 11 Apr 2011 10:07:52 GMT http://hdl.handle.net/2117/12330 2011-04-11T10:07:52Z Olier Caparroso, Iván; Vellido Alcacena, Alfredo no Multivariate time series, Generative topographic mapping, Unsupervised relevance determination, Clustering, Change point detection Most of the existing research on multivariate time series concerns supervised forecasting problems. In comparison, little research has been devoted to their exploration through unsupervised clustering and visualization. In this paper, the capabilities of Generative Topographic Mapping Through Time, a model with foundations in probability theory, that performs simultaneous time series clustering and visualization, are assessed in detail. Focus is placed on the visualization of the evolution of signal regimes and the exploration of sudden transitions, for which a novel identification index is defined. The interpretability of time series clustering results may become extremely difficult, even in exploratory visualization, for high dimensional datasets. Here, we define and test an unsupervised time series relevance determination method, fully integrated in the Generative Topographic Mapping Through Time model, that can be used as a basis for time series selection. This method should ease the interpretation of time series clustering results. http://hdl.handle.net/2117/10946 Title: Feature Selection with Single-Layer Perceptrons for a multicentre 1H-MRS brain tumour database Authors: Romero Merino, Enrique; Vellido Alcacena, Alfredo; Sopena, Josep Maria Abstract: 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. Mon, 10 Jan 2011 17:29:58 GMT http://hdl.handle.net/2117/10946 2011-01-10T17:29:58Z Romero Merino, Enrique; Vellido Alcacena, Alfredo; Sopena, Josep Maria no 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. http://hdl.handle.net/2117/10945 Title: Rule-based assistance to brain tumour diagnosis using LR-FIR Authors: Nebot Castells, M. Àngela; Castro Espinoza, Félix Agustín; Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Arús, Carles Abstract: This paper describes a process of rule-extraction from a multi-centre brain tumour database consisting of nuclear magnetic res- onance spectroscopic signals. The expert diagnosis of human brain tumours can benefit from computer-aided assistance, which has to be readily interpretable by clinicians. Interpretation can be achieved through rule extraction, which is here performed using the LR-FIR algorithm, a method based on fuzzy logic. The experimental results of the classiffication of three groups of tumours indicate in this study that just three spectral frequencies, out of the 195 from a range pre-selected by experts, are enough to represent, in a simple and intuitive manner, most of the knowledge required to discriminate these groups. Mon, 10 Jan 2011 17:22:59 GMT http://hdl.handle.net/2117/10945 2011-01-10T17:22:59Z Nebot Castells, M. Àngela; Castro Espinoza, Félix Agustín; Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Arús, Carles no Rule extraction, Fuzzy inductive reasoning, Brain tumours, Magnetic resonance spectroscopy, Medical decision support systems This paper describes a process of rule-extraction from a multi-centre brain tumour database consisting of nuclear magnetic res- onance spectroscopic signals. The expert diagnosis of human brain tumours can benefit from computer-aided assistance, which has to be readily interpretable by clinicians. Interpretation can be achieved through rule extraction, which is here performed using the LR-FIR algorithm, a method based on fuzzy logic. The experimental results of the classiffication of three groups of tumours indicate in this study that just three spectral frequencies, out of the 195 from a range pre-selected by experts, are enough to represent, in a simple and intuitive manner, most of the knowledge required to discriminate these groups. http://hdl.handle.net/2117/10944 Title: Exploratory characterization of outliers in a multi-centre 1H-MRS brain tumour dataset Authors: Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Romero Merino, Enrique; Arús, Carles Abstract: 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. Mon, 10 Jan 2011 17:13:47 GMT http://hdl.handle.net/2117/10944 2011-01-10T17:13:47Z Vellido Alcacena, Alfredo; Julià Sapé, Margarida; Romero Merino, Enrique; Arús, Carles no Medical decision support systems, Proton magnetic resonance spectroscopy, Brain tumours, Outlier detection, Data exploration, Data visualization, Dimensionality reduction 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. http://hdl.handle.net/2117/9513 Title: Geodesic Generative Topographic Mapping Authors: Cruz Barbosa, Raúl; Vellido Alcacena, Alfredo Abstract: Nonlinear dimensionality reduction (NLDR) methods aim to provide a faithful low-dimensional representation of multivariate data. The manifold learning family of NLDR methods, in particular, do this by defining low-dimensional manifolds embedded in the observed data space. Generative Topographic Mapping (GTM) is one such manifold learning method for multivariate data clustering and visualization. The non-linearity of the mapping it 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 study, the GTM is modified to prioritize neighbourhood relationships along the generated manifold. This is accomplished through penalizing 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 model is shown to improve not only the continuity and trustworthiness of the representation generated by the model, but also its resilience in the presence of noise. Thu, 07 Oct 2010 10:04:35 GMT http://hdl.handle.net/2117/9513 2010-10-07T10:04:35Z Cruz Barbosa, Raúl; Vellido Alcacena, Alfredo no Nonlinear dimensionality reduction (NLDR) methods aim to provide a faithful low-dimensional representation of multivariate data. The manifold learning family of NLDR methods, in particular, do this by defining low-dimensional manifolds embedded in the observed data space. Generative Topographic Mapping (GTM) is one such manifold learning method for multivariate data clustering and visualization. The non-linearity of the mapping it 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 study, the GTM is modified to prioritize neighbourhood relationships along the generated manifold. This is accomplished through penalizing 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 model is shown to improve not only the continuity and trustworthiness of the representation generated by the model, but also its resilience in the presence of noise. http://hdl.handle.net/2117/9511 Title: Unfolding the Manifold in Generative Topographic Mapping Authors: Cruz Barbosa, Raúl; Vellido Alcacena, Alfredo Abstract: Generative Topographic Mapping (GTM) is a probabilistic latent variable model for multivariate data clustering and visualization. It tries to capture the relevant data structure by defining a low-dimensional manifold embedded in the high-dimensional data space. This requires the assumption that the data can be faithfully represented by a manifold of much lower dimension than that of the observed space. Even when this assumption holds, the approximation of the data may, for some datasets, require plenty of folding, resulting in an entangled manifold and in breaches of topology preservation that would hamper data visualization and cluster definition. This can be partially avoided by modifying the GTM learning procedure so as to penalize divergences between the Euclidean distances from the data to the model prototypes and the corresponding geodesic distances along the manifold. We define and assess this strategy, comparing it to the performance of the standard GTM, using several artificial datasets. Thu, 07 Oct 2010 09:57:36 GMT http://hdl.handle.net/2117/9511 2010-10-07T09:57:36Z Cruz Barbosa, Raúl; Vellido Alcacena, Alfredo no Generative Topographic Mapping (GTM) is a probabilistic latent variable model for multivariate data clustering and visualization. It tries to capture the relevant data structure by defining a low-dimensional manifold embedded in the high-dimensional data space. This requires the assumption that the data can be faithfully represented by a manifold of much lower dimension than that of the observed space. Even when this assumption holds, the approximation of the data may, for some datasets, require plenty of folding, resulting in an entangled manifold and in breaches of topology preservation that would hamper data visualization and cluster definition. This can be partially avoided by modifying the GTM learning procedure so as to penalize divergences between the Euclidean distances from the data to the model prototypes and the corresponding geodesic distances along the manifold. We define and assess this strategy, comparing it to the performance of the standard GTM, using several artificial datasets. http://hdl.handle.net/2117/6880 Title: Data mining in cancer research Authors: Lisboa, Paulo J.G.; Vellido Alcacena, Alfredo; Tagliaferri, Roberto; Napolitano, Francesco; Ceccarelli, Michelle; Martín Guerrero, José D.; Biganzoli, Elia Abstract: This article is not intended as a comprehensive survey of data mining applications in cancer. Rather, it provides starting points for further, more targeted, literature searches, by embarking on a guided tour of computational intelligence applications in cancer medicine, structured in increasing order of the physical scales of biological processes. Thu, 08 Apr 2010 10:19:11 GMT http://hdl.handle.net/2117/6880 2010-04-08T10:19:11Z Lisboa, Paulo J.G.; Vellido Alcacena, Alfredo; Tagliaferri, Roberto; Napolitano, Francesco; Ceccarelli, Michelle; Martín Guerrero, José D.; Biganzoli, Elia no This article is not intended as a comprehensive survey of data mining applications in cancer. Rather, it provides starting points for further, more targeted, literature searches, by embarking on a guided tour of computational intelligence applications in cancer medicine, structured in increasing order of the physical scales of biological processes.