
handle: 10810/36584
Este proyecto se basa en el estudio de los algoritmos que conforman una red neuronal artificial. Para ello se estudia primero la teoría matemática que hay detrás, partiendo de los métodos clásicos de aproximación de funciones y evolucionando a su adaptación en las redes, mostrando especial atención al algoritmo de Backpropagation. Posteriormente se estudia su aplicación en Matlab, adaptándolos para sacar el máximo beneficio a las capacidades matriciales propias del programa y a la programación orientada a objeto, con que se desarrollan los contenidos de la red neuronal. Ver los algoritmos aplicados y su funcionamiento en los ejemplos permite comprenderlos mejor, así como poder observar las ventajas y desventajas de los distintos métodos. Español
The main objective of this project is the study of the algorithms that form an artificial neural network. To do that, we will study,first,the mathematical theory behind them, starting from the classical methods of function approximation and evolving to their adaptation to the networks, payingspecial attention to the Backpropagation.After that, we will study their application in Matlab, adapting them to get maximum benefit to the capacities of the program and getting to program a Neural Network object, which will enable us to solve simple examples. Seeing the algorithms applied and working may help to get a better understanding, and alsoenables to see the advantages and disadvantages of the different methods.
Proiektu hau oinarritzen da sare neuronal artifiziala osatzen duten algoritmoenestudioan. Horretarako, lehenik teoria matematikoa ikasten da, funtzio-hurbilketa metodo klasikotik hasiz eta sareetara bere egokitzera eboluzionatuz, Backpropagation arreta berezi erakusten.Geroago Matlab-en aplikazioa ikasten da, programaren ahalmenei onurahandienaateratzekomoldatuzetaSare Neuronala objektubat programatzea lortuz,adibideerrazeiaurre egiteko. Aplikatutako algoritmoak etaadibideetanfuntzionamendua ikustea, baimentzen du haiek hobeto ulertzea, baita ere metodo desberdinen abantailak eta desabantailak ikusi ahal izatea.
redes neuronales, Matlab
redes neuronales, Matlab
| selected citations These citations are derived from selected sources. This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | 0 | |
| popularity This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network. | Average | |
| influence This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically). | Average | |
| impulse This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network. | Average |
