Modeli za preporučivanje autorskih tekstualnih materijala bazirani na sadržaju te dubokom učenju

undergraduate thesis

• Title (english): Content-based recommender models for authored textual materials based on deep learning
• Author: Marin Stojsavljević
• Mentor: Mario Brčić (mentor)
• Committee member: Mario Brčić (predsjednik povjerenstva)
• Committee member: Agneza Krajna (član povjerenstva)
• Committee member: Juraj Dončević (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-01, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U današnje vrijeme sustavi za preporuku igraju vrlo važnu ulogu u svakodnevnom životu. Algoritmi strojnoga učenja čine okosnicu takvih sustava, a u posljednje vrijeme veliki zamah u industriji donijeli su algoritmi dubokog učenja. U ovome radu usporedit će se dvije arhitekture dubokog učenja - ručno rađena arhitektura višeslojnog perceptrona i duboka arhitektura s dva tornja. One su naučene nad trima različitim skupovima podataka čija će se kvaliteta usporediti pomoću spomenutih dubokih arhitektura. Također, ispitat će se i rezultati arhitekture dobivene iz automatskog sustava koja samostalno bira model i namješta njegove parametre.

Abstract (english)

Today, recommendation systems play very important role in our everyday lives. Machine learning algorithms form the backbone of such systems, and in recent times deep learning algorithms have brought great momentum to the industry. This paper will compare two deep learning architectures - a handmade multilayer perceptron architecture and a deep architecture with two towers. They are trained on three different data sets whose quality will be compared using those deep architectures. Also, the results of the architecture obtained from an automatic system that independently selects the model and sets its parameters will be examined.

• Keywords: duboko učenje; sustavi za preporuku; višeslojni perceptron
• Keywords (english): deep learning; recommendation systems; multilayer perceptron
• Language: croatian
• URN:NBN: urn:nbn:hr:168:866277
• 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: 2023-01-14 21:41:58