Geometrijski pristup dubokom učenju

undergraduate thesis

• Title (english): Geometrical approach to deep learning
• Author: Mislav Đomlija
• Mentor: Dario Bojanjac (mentor)
• Committee member: Dario Bojanjac (predsjednik povjerenstva)
• Committee member: Filip Turčinović (član povjerenstva)
• Committee member: Dubravko Tomić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-06-29, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U ovom radu smo na jasan način prezentirali osnove Riemannove geometrije i njenu primjenu na standardnu arhitekturu neuronske mreže. Izveli smo način za računanje nivo hiperploha u podatkovnoj domeni i to primjenili na jednostavnu neuronsku mrežu koja obavlja logističku regresiju nad trodimenzionalnim točkama

Abstract (english)

In this paper, we presented the basics of Riemannian geometry and its application on standard neural network architecture. We developed a framework to calculate level hyperplanes in the data domain and we applied that to simple neural network, trained to do logistic regression on 3D points.

• Keywords: duboko učenje; mnogostrukost; Riemannova metrika; neuronska mrež; geometrijsko duboko učenje
• Keywords (english): deep learning; manifold; Riemann metric; neural network; geometric deep learning
• Language: croatian
• URN:NBN: urn:nbn:hr:168:667170
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-07-13 10:02:36