Introducción al machine learning con TensorFlow
Loading...
Identifiers
Publication date
Authors
Advisors
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
En este trabajo se estudian las bases del aprendizaje automático o Machine Learning centrándose en el aprendizaje automático supervisado, en el que se le proporcionan a la máquina un conjunto de datos etiquetados. Se detalla el funcionamiento de la clasificación como técnica para estudiar los datos de entrada, que se basa en el uso de redes neuronales artificiales. En concreto, a lo largo del trabajo se analizan las redes neuronales artificiales convolucionales que son de gran utilidad a la hora de procesar datos de tipo cuadrícula, como las imágenes, para ello hacen uso de un optimizador de la función de coste del problema. En este caso estudiaremos las ventajas y el funcionamiento del método de descenso Adam. Finalmente, con el objetivo de poner en práctica todos los conceptos analizados a lo largo del trabajo, se entrena un algoritmo dedicado a la clasificación de imágenes, empleando para ello el conjunto de datos CIFAR-10 y la librería Keras de Tensorflow
This paper examines the basics of machine learning focusing on supervised machine learning, in which a set of tagged data is provided to the machine. It details the functioning of the classification as a technique to study input data, which is based on the use of artificial neural networks. Specifically, throughout the project, arti cial convolutional neural networks are analyzed, as they are very useful when processing grid-type data, such as images, for this purpose we employ an optimizer of the cost function of the problem, and in particular we will study the advantages and functioning of the Adam descent method. Finally, with the aim of putting into practice all the concepts analyzed throughout the project, an algorithm dedicated to the classification of images is trained using the CIFAR-10 data set and the Keras library of Tensorflow
This paper examines the basics of machine learning focusing on supervised machine learning, in which a set of tagged data is provided to the machine. It details the functioning of the classification as a technique to study input data, which is based on the use of artificial neural networks. Specifically, throughout the project, arti cial convolutional neural networks are analyzed, as they are very useful when processing grid-type data, such as images, for this purpose we employ an optimizer of the cost function of the problem, and in particular we will study the advantages and functioning of the Adam descent method. Finally, with the aim of putting into practice all the concepts analyzed throughout the project, an algorithm dedicated to the classification of images is trained using the CIFAR-10 data set and the Keras library of Tensorflow
Description
Traballo Fin de Grao en Matemáticas. Curso 2021-2022
Keywords
Bibliographic citation
Relation
Has part
Has version
Is based on
Is part of
Is referenced by
Is version of
Requires
Sponsors
Rights
Atribución-NoComercial-CompartirIgual 4.0 Internacional







