Shortened Persistent Homology for a Biomedical Retrieval System with Relevance Feedback - Machine Learning and Knowledge Extraction Access content directly
Conference Papers Year : 2018

Shortened Persistent Homology for a Biomedical Retrieval System with Relevance Feedback

Abstract

This is the report of a preliminary study, in which a new coding of persistence diagrams and two relevance feedback methods, designed for use with persistent homology, are combined. The coding consists in substituting persistence diagrams with complex polynomials; these are “shortened”, in the sense that only the first few coefficients are used. The relevance feedback methods play on the user’s feedback for changing the impact of the different filtering functions in determining the output.
Fichier principal
Vignette du fichier
472936_1_En_20_Chapter.pdf (415.42 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02060046 , version 1 (07-03-2019)

Licence

Attribution

Identifiers

Cite

Alessia Angeli, Massimo Ferri, Eleonora Monti, Ivan Tomba. Shortened Persistent Homology for a Biomedical Retrieval System with Relevance Feedback. 2nd International Cross-Domain Conference for Machine Learning and Knowledge Extraction (CD-MAKE), Aug 2018, Hamburg, Germany. pp.282-292, ⟨10.1007/978-3-319-99740-7_20⟩. ⟨hal-02060046⟩
46 View
54 Download

Altmetric

Share

Gmail Facebook X LinkedIn More