Now showing items 21-40 of 42

    • 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 ...
    • Light machine vision camera platform with MIPI interface for on-board AI systems 

      Ansola Erkizia, Julen (Universitat Politècnica de Catalunya, 2019-09-02)
      Master thesis
      Restricted access - confidentiality agreement
      Interfaces for artificial intelligence systems like GigE or PoCL lead to big dimensions and relatively high power consumptions. MIPI has released a scalable interface for image capturing, named CSI, that can reduce these ...
    • Making kernel machines scalable combining matrix approximations and distributed computing 

      Gerstenlauer, Jakob L.K. (Universitat Politècnica de Catalunya, 2018-04-08)
      Master thesis
      Open Access
      In this work, kernelized binary support vector machines are implemented based on stochastic gradient descent. The Scala library can be used both on a single computing node and on a Spark cluster. Additional tools for ...
    • Microbiological data analysis with non-linear methods 

      Sannidhi, Nikhith (Universitat Politècnica de Catalunya, 2018-06-25)
      Bachelor thesis
      Open Access
      Aquest projecte estudiarà l'aplicació de l'estat estadístic de la tècnica i els mètodes d'aprenentatge automàtic per desenvolupar models predictius i determinar paràmetres de discriminació òptims per al seguiment d'origen ...
    • Multivariate Kernel Functions for Categorical Variables 

      Garcia Marquez, Carlos (Universitat Politècnica de Catalunya, 2014-12-16)
      Master thesis (pre-Bologna period)
      Open Access
      Recerca sobre kernels aplicables a variables de tipus categoric, basades en l'ús de la funció de probabilitat de les dades. Introdueix noves funcions kernel inspirades en mesures de semblança per a variables binàries.
    • Network Function Virtualization technologies applied to cellular systems 

      Padilla Pinedo, Joel (Universitat Politècnica de Catalunya, 2020-02)
      Master thesis
      Open Access
      Future 5G networks open a new spectrum of possibilities, both at the level of services it can offer, and at the level of its deployment. This thesis aims to make a study of some of the technologies that make possible the ...
    • New hybrid kernel architectures for deep learning 

      Mora de Checa, Daniel (Universitat Politècnica de Catalunya, 2018-04)
      Master thesis
      Open Access
      In this work we explore the possibilities of combining neural network architectures and kernel methods by introducing hybrid kernel blocks. We present hybrid architectures which can be trained as traditional neural networks ...
    • Optimization of OpenCL applications on FPGA 

      Navarro Torrentó, Albert (Universitat Politècnica de Catalunya, 2018-04)
      Master thesis
      Open Access
      Covenantee:   Barcelona Supercomputing Center
      This document presents an evaluation of OpenCL as a mechanism to exploit FPGA resources. To evaluate it, we show a performance and energy comparison between an Intel Arria 10 and an Intel Xeon E5-2600. We also present a ...
    • Optimizing support vector machines (SVMs) with optimized geometric ensembles (OFE) 

      Guitart Atienza, Anna (Universitat Politècnica de Catalunya, 2015-04-30)
      Master thesis
      Open Access
      Delivery of my master thesis with the new title: Optimizing Support Vector Machines with Characteristic Boundary Points.
    • Optimizing the SpMV kernel on long-vector accelerators 

      Gómez Crespo, Constantino; Mantovani, Filippo; Focht, Erich; Casas, Marc (Barcelona Supercomputing Center, 2021-05)
      Conference report
      Open Access
      Sparse Matrix-Vector multiplication (SpMV) is an essential kernel for parallel numerical applications. SpMV displays sparse and irregular data accesses, which complicate its vectorization. Such difficulties make SpMV ...
    • Performing flexible control on low cost microcontrollers using a minimal real-time kernel 

      Marau, Ricardo; Leite, Paulo; Velasco García, Manel; Martí Colom, Pau; Almeida, Luis; Pedreiras, Paulo; Fuertes Armengol, José Mª (2008-05-01)
      Article
      Open Access
      In recent years, approaches to control performance and resource optimization for embedded control systems have been receiving increased attention. Most of them focus on theory, whereas practical aspects are omitted. ...
    • Predictive techniques to support the adjudication of construction works 

      Lahuerta Piñeiro, Guillermo (Universitat Politècnica de Catalunya, 2020-07-01)
      Master thesis
      Restricted access - confidentiality agreement
    • Selection of basis functions guided by the L2 soft margin 

      Barrio Moliner, Ignacio; Romero Merino, Enrique; Belanche Muñoz, Luis Antonio (Springer, 2007)
      Conference report
      Open Access
      Support Vector Machines (SVMs) for classification tasks produce sparse models by maximizing the margin. Two limitations of this technique are considered in this work: firstly, the number of support vectors can be large ...
    • Sentiment analysis of abusive texts 

      González Benitez, Esther (Universitat Politècnica de Catalunya, 2019-06)
      Bachelor thesis
      Restricted access - confidentiality agreement
    • Soporte para depuración en FPGA 

      Meléndez Núñez, Christian (Universitat Politècnica de Catalunya, 2018-04-04)
      Bachelor thesis
      Open Access
      Describiremos la implementación de un soporte a la depuración llamado hgdb basado en la idea que sigue LegUp, un entorno de depuración orientado a la síntesis de alto nivel (HLS) que permite inspeccionar circuitos generados ...
    • Sparse methods for Online Support Vector Machine 

      Riera Molina, Carles Roger (Universitat Politècnica de Catalunya, 2014-11-28)
      Master thesis
      Open Access
      Aquest treball proposa un nou classificador online basat en Kernel SVM que minimitza el nombre de vectors de suport gràcies una restricció adicional imposada mitllançant métodes de continuació. Conseqüentment, es pot fer ...
    • Study and extension of kernel functions for missing value treatment 

      Hernández Potiomkin, Yaroslav (Universitat Politècnica de Catalunya, 2015-01-20)
      Bachelor thesis
      Open Access
    • The effect of statistical normalisation on network propagation scores 

      Picart Armada, Sergio; Thompson, Wesley; Buil Demur, Alfonso; Perera Lluna, Alexandre (2020-10-18)
      Article
      Open Access
      Network diffusion and label propagation are fundamental tools in computational biology, with applications like gene-disease association, protein function prediction and module discovery. More recently, several publications ...
    • Towards fast hybrid deep kernel learning methods 

      Lara Miquel, Miquel (Universitat Politècnica de Catalunya, 2019-04-15)
      Master thesis
      Open Access
      El treball estudia la millor manera de crear xarxes neuronals híbrides amb mètodes kernel mitjançant dues aproximacions de kernel diferents, random Fourier features i el mètode Nystrom, i la millor manera d'entrenar-les, ...
    • Transformer-based kernels for Natural Language Processing 

      Morán Pomés, David (Universitat Politècnica de Catalunya, 2020-10-22)
      Master thesis
      Open Access
      Kernel methods (KM) and Artificial Neural Networks (ANN) are two of the main families of methods that are more widely spread nowadays in the world of Machine Learning, being the Support Vector Machine (SVM) and the Deep ...