Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches

Διπλωματική εργασία--Πανεπιστήμιο Μακεδονίας, Θεσσαλονίκη, 2024.

Πρώτος συγγραφέας: Πασβάντης, Κωνσταντίνος
Επόπτης Καθηγητής: Πρωτοπαπαδάκης, Ευτύχιος
Μορφή: Electronic Thesis or Dissertation
Γλώσσα: English
Άλλες Λεπτομέρειες Έκδοσης: Πανεπιστήμιο Μακεδονίας, 2024
Τμήμα: Πρόγραμμα Μεταπτυχιακών Σπουδών στην Τεχνητή Νοημοσύνη και Αναλυτική Δεδομένων
Θέματα/Λέξεις Κλειδιά:
Computer Vision
Health Informatics
Διαθέσιμο Online: http://dspace.lib.uom.gr/handle/2159/30853
id dspace-2159-30853
recordtype dspace
spelling dspace-2159-308532024-07-05T00:04:23Z Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches Πασβάντης, Κωνσταντίνος Πρωτοπαπαδάκης, Ευτύχιος Πρόγραμμα Μεταπτυχιακών Σπουδών στην Τεχνητή Νοημοσύνη και Αναλυτική Δεδομένων Computer Vision Health Informatics The application of deep learning models in medical diagnosis has showcased considerable efficacy in recent years. Nevertheless, a notable limitation involves the inherent lack of explainability during decision-making processes. This study addresses such a constraint, by enhancing the interpretability robustness. The primary focus is directed towards refining the explanations generated by the LIME Library and LIME image explainer. This is achieved throuhg post-processing mechanisms, based on scenario-specific rules. Multiple experiments have been conducted using publicly accessible datasets related to brain tumor detection. Our proposed post-heuristic approach demonstrates significant advancements, yielding more robust and concrete results, in the context of medical diagnosis. Διπλωματική εργασία--Πανεπιστήμιο Μακεδονίας, Θεσσαλονίκη, 2024. 2024-07-04T09:42:54Z 2024-07-04T09:42:54Z 2024-07-04T09:42:54Z 2024 Electronic Thesis or Dissertation http://dspace.lib.uom.gr/handle/2159/30853 en http://creativecommons.org/licenses/by-sa/4.0/ Αναφορά Δημιουργού - Παρόμοια Διανομή 4.0 Διεθνές Πανεπιστήμιο Μακεδονίας
institution University of Macedonia
collection DSpace collection
language English
topic Computer Vision
Health Informatics
spellingShingle Computer Vision
Health Informatics
Πασβάντης, Κωνσταντίνος
Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches
description Διπλωματική εργασία--Πανεπιστήμιο Μακεδονίας, Θεσσαλονίκη, 2024.
abstract The application of deep learning models in medical diagnosis has showcased considerable efficacy in recent years. Nevertheless, a notable limitation involves the inherent lack of explainability during decision-making processes. This study addresses such a constraint, by enhancing the interpretability robustness. The primary focus is directed towards refining the explanations generated by the LIME Library and LIME image explainer. This is achieved throuhg post-processing mechanisms, based on scenario-specific rules. Multiple experiments have been conducted using publicly accessible datasets related to brain tumor detection. Our proposed post-heuristic approach demonstrates significant advancements, yielding more robust and concrete results, in the context of medical diagnosis.
advisor Πρωτοπαπαδάκης, Ευτύχιος
format Electronic Thesis or Dissertation
author Πασβάντης, Κωνσταντίνος
author-letter Πασβάντης, Κωνσταντίνος
department Πρόγραμμα Μεταπτυχιακών Σπουδών στην Τεχνητή Νοημοσύνη και Αναλυτική Δεδομένων
title Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches
title_short Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches
title_full Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches
title_fullStr Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches
title_full_unstemmed Enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches
title_sort enhancing deep learning model explainability in brain tumor datasets using post-heuristic approaches
publisher Πανεπιστήμιο Μακεδονίας
publishDate 2024
url http://dspace.lib.uom.gr/handle/2159/30853
_version_ 1828238902378364928
score 13,073021

Παρόμοια τεκμήρια