Master in Innovation and Research in Informatics - MIRI
Enviaments recents
-
Performance analysis of BigData Frameworks using BigBench
(Universitat Politécnica de Catalunya, 2018-04)
Projecte Final de Màster Oficial
Accés obertThe ALOJA benchmarking platform is expanded with the first standardized Big Data benchmark: BigBench, in which a elasticity test extension is added. By making use of ALOJA and BigBench we compare the de facto SQL Big Data ... -
Simulating the Behavior of the Human Brain on NVIDIA GPU: cuHinesBatch & cuThomasBatch implementations
(Universitat Politécnica de Catalunya, 2018-01-15)
Projecte Final de Màster Oficial
Accés obertUnderstand the human brain is one of the century challenges. On this work we are going to achieve a small step towards this objective presenting a novel data layout in order to compute more efficiently the Hines algorithm ... -
Instance Segmentation of Point Clouds using Deep Learning
(Universitat Politécnica de Catalunya, 2018-01)
Projecte Final de Màster Oficial
Accés obertThe main objective of this thesis is to implement a deep network architecture to segment and instantiate objects on 3d point clouds using pointnet, a novel 3d deep learning network that segments 3d point clouds in semantically ... -
An approach to quantifying hardware diversity against common cause failures
(Universitat Politécnica de Catalunya / Universitat Politècnica de Catalunya, 2018-01-14)
Projecte Final de Màster Oficial
Accés obertIn this thesis, we cover the gapof quantifying diversity by introducing DIMP, a low-cost diversity metric based on analyzing the paths of the circuits and relating it to the particular case of automotive microcontrollers ... -
Design and implementation of gamified attention test for secondary school students
(Universitat Politécnica de Catalunya, 2017-10-24)
Projecte Final de Màster Oficial
Accés restringit per acord de confidencialitatEsta tesis busca introducir el diseño de un juego de aprendizaje para facilitar el proceso de aprendizaje en los estudiantes de educación secundaria. El diseño intenta aumentar la motivación para que el jugador se enganche ... -
Deep Learning for Protein Function Prediction
(Universitat Politècnica de Catalunya, 2017-09-26)
Projecte Final de Màster Oficial
Accés obert
Realitzat a/amb: Technische Universität MünchenThe assignment of functions to proteins is a bottleneck due to the need of costly and time-consuming molecular experiments. This is the reason why more often data analysis methods are used for protein an- notation. In this ... -
Semantic stream computing for large data-set analytics
(Universitat Politécnica de Catalunya, 2017-12)
Projecte Final de Màster Oficial
Accés obertOne of the most interesting research fields in IT industries is the quest of software able to deal efficiently with Big Data. In this thesis, we have studied Big Data especially about their spread in healthcare, where data ... -
Code generation for the dataflow-based XSMLL runtime
(Universitat Politécnica de Catalunya, 2017-10)
Projecte Final de Màster Oficial
Accés obert
Realitzat a/amb: Barcelona Supercomputing Centre (BSC)This document presents the report of the Master’s Thesis Code generation for the dataflow-based xsmll runtime developed in the context of the Master in Innovation and Research in Informatics. In this report we analyze ... -
Synchronization / communication techniques for OmpSs@FPGA
(Universitat Politécnica de Catalunya, 2017-10-24)
Projecte Final de Màster Oficial
Accés obert
Realitzat a/amb: Barcelona Supercomputing Centre (BSC)HPC machines are introducing more and more heterogeneity in their architecture on the road to exascale systems. The increasing complexity of the machines due to the variety of hardware architectures and accelerators makes ... -
Non-Conventional Vector Units for Big Data Workloads
(Universitat Politècnica de Catalunya, 2017-10-16)
Projecte Final de Màster Oficial
Accés restringit per acord de confidencialitatThis work presents the improvements applied to vector units, to enable them to find and properly use Data Level Parallelism in applications with index matching algorithms and less regular data access patterns. Our designs ... -
Pipeline design to identify key features and perform classification on response/predisposition large-scale genetic data
(Universitat Politècnica de Catalunya, 2017-04)
Projecte Final de Màster Oficial
Accés obertWe propose a state-of-the-art, scalable and flexible alternative to the classical GWAS approach, based on machine learning techniques, to analyze large-scale data and discover epistatic and non-epistatic polygenic variants ... -
Claims classification in an automative company
(Universitat Politècnica de Catalunya, 2017-05)
Projecte Final de Màster Oficial
Accés restringit per acord de confidencialitat
RSS 1.0