Graph Convolutional Neural Networks applied to classify cancer types

Identifier:  TFM:992

Handle: https://hdl.handle.net/20.500.11797/TFM992

Authors:  Pascual Saldaña, Heribert

Abstract:
The main objective of this master's thesis was to reproduce the results published in the paper “Classification of Cancer Types Using Graph Convolutional Neural Networks” written by Ricardo Ramirez et al., since the use of these networks is a relatively new concept. The paper’s goal is to classify tumour tissues correctly using their RNA, through the use of convolutional graph networks (aka. GCNN). Finally, the code supplied with the paper isn’t usable, because there’re some missing parts and is focused in show the results exposed in the paper. For these reasons, many modifications and proposals are shown in this thesis.