Neural networks for melanoma diagnosis

Paper Title:

Are ensembles getting better all the time?

Published on:

29 November 2023

Primary Category:

Machine Learning

Paper Authors:

Pierre-Alexandre Mattei,

Damien Garreau

Bullets

Key Details

•

Trains convolutional neural network for melanoma diagnosis

•

Uses dermatoscopic image dataset of benign and malignant lesions

•

Neural nets effective for melanoma ID, ensembles improve accuracy

•

Cover image: Neural net visualizations of learned melanoma features

Explore the topics in this paper

convolutional neural networks

deep learning

medical imaging

melanoma detection

skin cancer diagnosis

AI generated summary

Neural networks for melanoma diagnosis

This paper trains a convolutional neural network to diagnose melanoma skin cancer from dermatoscopic images. The network is trained on a dataset of benign and malignant skin lesions. Key results show neural nets can effectively identify melanoma, though ensembles improve accuracy. The cover image concept depicts neural network visualization techniques highlighting discriminative melanoma features learned by the model.