Real-time scalable parallel data stream classification

The main objective of this final master project is to create a real-time prototype that is capable of classifying real-time data using several deep learning algorithms. Classifying means to give "valuable" information ¿ that maybe can be unknown - to the different incoming data. Note also...

ver descrição completa

Detalhes bibliográficos
Autor: Robledo Mcclymont, Roberto Dean
Formato: tesis de maestría
Fecha de publicación:2018
País:España
Recursos:Universitat Oberta de Catalunya (UOC)
Repositorio:O2, repositorio institucional de la UOC
OAI Identifier:oai:openaccess.uoc.edu:10609/83465
Acesso em linha:http://hdl.handle.net/10609/83465
Access Level:acceso abierto
Palavra-chave:big data
High Performance Computing
artificial intelligence
Apache Kafka
deep learning
aprenentatge profund
intel·ligència artificial
computació d'alt rendiment
inteligencia artificial
aprendizaje profundo
computación de alto rendimiento
Big data -- TFM
Dades massives -- TFM
Datos masivos -- TFM
Descrição
Resumo:The main objective of this final master project is to create a real-time prototype that is capable of classifying real-time data using several deep learning algorithms. Classifying means to give "valuable" information ¿ that maybe can be unknown - to the different incoming data. Note also that this could be extrapolated to other fields. In addition, some research will be done in the field of deep learning with the aim of giving some guidelines about how big data can be classified in a cluster environment. The idea of developing this prototype is to prove that large amounts of data processing can be tackled within this methodology. Further work can be done following this line with the purpose of creating a real data-time analysis methodology that can be applicable to other fields such us medical studies, economic statistics, mobility solutions and many others. As in all research studies, iterative processing must be done in order to enhance and/or update the deep algorithms that will be presented during this final master project.