Use this link to cite:
http://hdl.handle.net/2183/33905 Selección de características federada basada en información mutua
Loading...
Identifiers
Publication date
Authors
González Fraga, Iván
Other responsabilities
Universidade da Coruña. Facultade de Informática
Journal Title
Bibliographic citation
Type of academic work
Academic degree
Abstract
[Resumen]: El crecimiento constante de los dispositivos del Internet de las cosas (IoT) ha llevado a problemas
de escalabilidad, privacidad y seguridad. La enorme cantidad de datos generados por estos
dispositivos plantea desafíos en términos de manejo y protección de información sensible. La
técnica de aprendizaje federado surge como una solución, permitiendo entrenar modelos de
inteligencia artificial en los propios dispositivos sin necesidad de transferir los datos, preservando
así la privacidad. En este proyecto se desarrollará y aplicará un enfoque de aprendizaje
federado, utilizando la medida de Información Mutua (IM) como base para un método de selección
de características, la cual es un paso de preprocesado importante para reducir la alta
dimensionalidad de los datos.
[Abstract]: The constant growth of Internet of Things (IoT) devices has led to scalability, privacy, and security issues. The enormous amount of data generated by these devices poses challenges in terms of handling and protecting sensitive information. The federated learning technique emerges as a solution, allowing artificial intelligence models to be trained on the devices themselves without the need to transfer the data, thus preserving privacy. In this project, a federated learning approach will be developed and applied, using the Mutual Information (MI) measure as the basis for a feature selection method, which is an important preprocessing step to reduce high-dimensionality of data.
[Abstract]: The constant growth of Internet of Things (IoT) devices has led to scalability, privacy, and security issues. The enormous amount of data generated by these devices poses challenges in terms of handling and protecting sensitive information. The federated learning technique emerges as a solution, allowing artificial intelligence models to be trained on the devices themselves without the need to transfer the data, thus preserving privacy. In this project, a federated learning approach will be developed and applied, using the Mutual Information (MI) measure as the basis for a feature selection method, which is an important preprocessing step to reduce high-dimensionality of data.
Description
Editor version
Rights
Todos los derechos reservados.







