Adversary detection in neural networks via persistent homology

Belchí Guillamón, FranciscoDa Dalt, Severino2021-11-162021-11-162021-01-24https://hdl.handle.net/2445/181290Treballs Finals de Grau de Matemàtiques, Facultat de Matemàtiques, Universitat de Barcelona, Any: 2021, Director: Francisco Belchí Guillamón[en] The main goal of this work is to present a recently-invented homology theory called persistent homology and its application on the detection of adversary examples of neural network presented in the paper [4].35 p.application/pdfengcc-by-nc-nd (c) Severino Da Dalt, 2021http://creativecommons.org/licenses/by-nc-nd/3.0/es/HomologiaTreballs de fi de grauXarxes neuronals (Informàtica)Aprenentatge automàticTopologia algebraicaHomologyBachelor's thesesNeural networks (Computer science)Machine learningAlgebraic topologyAdversary detection in neural networks via persistent homologyinfo:eu-repo/semantics/bachelorThesisinfo:eu-repo/semantics/openAccess