Now showing items 1-20 of 41

    • A probabilistic tri-class Support Vector Machine 

      González Abril, Luis; Angulo Bahón, Cecilio; Velasco Morente, Francisco; Ortega Ramírez, Juan Antonio (2010-07)
      Article
      Open Access
      A probabilistic interpretation for the output obtained from a tri-class Support Vector Machine into a multi-classification problem is presented in this paper. Probabilistic outputs are defined when solving a multi-class ...
    • A study on output normalization in multiclass SVMs 

      González Abril, Luis; Velasco Morente, Francisco; Angulo Bahón, Cecilio; Ortega Ramírez, Juan Antonio (2013-02-01)
      Article
      Restricted access - publisher's policy
      The use of binary support vector machines (SVMs) in multi-classification is addressed in this paper. Margins associated to the bi-classifiers, since they depend on the geometrical disposition of the classes being separated, ...
    • An Android real-time kernel and system interface for open nano-satellite constellations 

      Marí Barceló, Marc (Universitat Politècnica de Catalunya, 2016-06)
      Bachelor thesis
      Open Access
      L'objectiu d'aquest treball és dissenyar i implementar part de l'arquitectura de software per a una plataforma de desenvolupament de nano-satèl·lits oberta basada en Android. Per un costat, afegir temps real al kernel. Per ...
    • An investigation into new kernels for categorical variables 

      Villegas García, Marco Antonio (Universitat Politècnica de Catalunya, 2013-01)
      Master thesis
      Open Access
      Kernel-based methods first appeared in the form of support vector machines. Since the first Support Vector Machine (SVM) formulation in 1995, we have seen how the number of proposed kernel functions has quickly grown, ...
    • Analysis of kernel matrices and their relation to SVM performance 

      Wiejacha, Malgorzata (Universitat Politècnica de Catalunya, 2018-06-18)
      Master thesis
      Open Access
      The thesis explores the relationship between kernel matrix properties and RVM model performance in terms of its generalization power. It pinpoints desirable qualities of a kernel matrix and provides a heuristic for kernel ...
    • Análisis de diferentes métodos de Neural style transfer aplicados a vídeos 

      Camacho Gómez, Alberto (Universitat Politècnica de Catalunya, 2021-01)
      Bachelor thesis
      Open Access
      Los métodos de Neural style transfer consisten en extraer el estilo artístico de una imagen y aplicar este mismo estilo a otra imagen cualquiera. Recientemente se está innovando bastante en este campo, creando diferentes ...
    • Averaging of kernel functions 

      Belanche Muñoz, Luis Antonio; Tosi, Alessandra (2012)
      Conference report
      Open Access
      In kernel-based machines, the integration of several kernels to build more flexible learning methods is a promising avenue for research. In particular, in Multiple Kernel Learning a compound kernel is build by learning a ...
    • Data analytics for performance evaluation under uncertainties applied to an industrial refrigeration plant 

      Cirera Balcells, Josep; Cariño Corrales, Jesús Adolfo; Zurita Millán, Daniel; Ortega Redondo, Juan Antonio (Institute of Electrical and Electronics Engineers (IEEE), 2019-01-01)
      Article
      Open Access
      Artificial intelligence has bounced into industrial applications contributing several advantages to the field and have led to the possibility to open new ways to solve many actual problems. In this paper, a data-driven ...
    • Detection of dependence based on kernel signal processing 

      Cabrera Estanyol, Ferran de (Universitat Politècnica de Catalunya, 2017-02)
      Master thesis
      Open Access
    • Developments in kernel design 

      Belanche Muñoz, Luis Antonio (2013)
      Conference report
      Open Access
      The aim of this paper is to give a concise overview of kernels, with a special attention to non-standard or heterogeneous data sources (e.g. non-numerical or structured data). A second goal is to discuss the world of ...
    • Filtratge de dades per a Support Vector Machines 

      Morán Pomés, David (Universitat Politècnica de Catalunya, 2018-01-22)
      Bachelor thesis
      Open Access
      En aquest treball es presenta un nou algoritme anomenat "Filtratge per Puresa", dissenyat per a realitzar un filtratge de les dades d'entrenament d'una SVM per a classificació binària. Els experiments mostren reduccions ...
    • Forecasting financial time series using multiple Kernel Learning 

      Fábregues de los Santos, Luis (Universitat Politècnica de Catalunya, 2017-07-05)
      Master thesis
      Open Access
      This thesis introduces a forecasting procedure based on Multiple Kernel Learning to predict and measure the influence of several economic variables in the process of predicting the equity premium of the S&P 500 Index. ...
    • Geomasking: Análisis y comparativa de algoritmos para perturbaciones de puntos. 

      Gómez Calero, Raúl (Universitat Politècnica de Catalunya, 2018-01)
      Bachelor thesis
      Open Access
      Se ha propuesto como objetivo, partiendo de una investigación previa teórica, detectar la existencia de una relación directa y/o indirecta entre el grado de enmascaramiento de datos geográficos y los resultados obtenidos ...
    • Implementació i estudi d'una deep/wide kernel machine 

      Serra Alza, Xavier (Universitat Politècnica de Catalunya, 2015-06-23)
      Bachelor thesis
      Open Access
      Aquest projecte busca implementar i provar un tipus particular de Deep Kernel Machine. Aquest tipus de programa pertany al camp del Machine Learning i, per tant, s'usa per a fer predicions basades en un procés previ ...
    • Implementación y estudio de una wide kernel machine basada en RBMs 

      Dávila Carmona, Cristian (Universitat Politècnica de Catalunya, 2015-06)
      Bachelor thesis
      Open Access
      En este proyecto se estudia la utilización de una Restricted Boltzmann Machine como extractor de características de imágenes que se utilizarán para mejorar la clasificación que realizan dos métodos de aprendizaje estadístico: ...
    • Improving sparsity in online kernel models 

      Orihuela Salvatierra, Helena (Universitat Politècnica de Catalunya, 2013-09-14)
      Master thesis
      Open Access
      In this thesis, background theory about the online kernel-based algorithms and their use for online learning is presented. The analysis of the state-ofthe- art methods highlights an important drawback in many kernel ...
    • Integration of GPU kernels with deel learning frameworks 

      Boza i Dieguez, August (Universitat Politècnica de Catalunya, 2019-10-22)
      Bachelor thesis
      Open Access
      Les xarxes neuronals han experimentat un creixement exponencial en els darrers anys. Han crescut en complexitat i mida, però tot i que hi ha hagut millores en el rendiment encara hi ha molt per fer. En aquest projecte ...
    • Kernel alignment for identifying objective criteria from brain MEG recordings in schizophrenia 

      Martín Muñoz, Mario; Béjar Alonso, Javier; Espósito, Gennaro; Catala Roig, Neus; Cortés García, Claudio Ulises; Viñas, Ferran; Tarragó Bofarull, Josep; Rojo, Emilio; Nowak, Rafal (Elsevier, 2016-08-02)
      Article
      Open Access
      The current wide access to data from different neuroimaging techniques has permitted to obtain data to explore the possibility of finding objective criteria that can be used for diagnostic purposes. In order to decide which ...
    • Kernel and user level execution trace analysis for multi-core distributed Linux systems 

      Dagenais, Michel (Barcelona Supercomputing Center, 2016-09-10)
      Open Access
      The presentation will briefly introduce the organization of the DORSAL laboratory and the collaborative research and development projects undertaken with the financial support of industrial partners, including Ericsson, ...
    • Kernel functions for categorical variables with application to problems in the life sciences 

      Belanche Muñoz, Luis Antonio; Villegas, Marco (2013)
      Conference report
      Open Access
      We introduce a family of positive definite kernels specifically designed for problems described by categorical information. The kernels are based on the comparison of the probability mass function of the variables and have ...