Layer Embedding Analysis in Convolutional Neural Networks for Improved Probability Calibration and Classification

In this project, our goal is to develop a method for interpreting how a neural network makes layer-by-layer embedded decisions when trained for a classification task, and also to use this insight for improving the model performance. To do this, we first approximate the distribution of the image repr...

Descripción completa

Detalles Bibliográficos
Autores principales: Zhang, Fan, Dvornek, Nicha, Yang, Junlin, Chapiro, Julius, Duncan, James
Formato: Online Artículo Texto
Lenguaje:English
Publicado: 2020
Materias:
Article
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7606489/
https://www.ncbi.nlm.nih.gov/pubmed/32356739
http://dx.doi.org/10.1109/TMI.2020.2990625

Internet

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7606489/
https://www.ncbi.nlm.nih.gov/pubmed/32356739
http://dx.doi.org/10.1109/TMI.2020.2990625

Ejemplares similares