Algoritmi strojnog učenja za predviđanje gledanosti televizijskih kanala

master's thesis

• Title (english): Machine Learning Algorithms for Forecasting Television Channels Viewership
• Author: Marko Josipović
• Mentor: Alan Jović (mentor)
• Committee member: Alan Jović (predsjednik povjerenstva)
• Committee member: Jan Šnajder (član povjerenstva)
• Committee member: Marko Čupić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-09-17, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad opisuje teorijsku podlogu algoritama strojnog učenja koji su razmatrani za predviđanje gledanosti televizijskih kanala. Eksperimentalna usporedba odabranih algoritama pokazala je da algoritam stroja potpornih vektora za regresiju (SVR) daje najbolje rezultate predviđanja dostupnih podataka o stvarnim gledanostima. Algoritam linearne regresije ne daje zadovoljavajuće rezultate, a algoritam slučajne šume (ExtraTrees) ima malo lošije rezultate od stroja potpornih vektora. Opisana je i implementacija web aplikacije koja korisnicima omogućava predviđanje gledanosti na temelju učitanog rasporeda emitiranja. Aplikacija se može koristiti kao alat za optimizaciju rasporeda. Zbog velike vremenske složenosti algoritma stroja potpornih vektora, za korištenje u aplikaciji odabran je algoritam slučajne šume.

Abstract (english)

This paper describes the theoretical basis of machine learning algorithms that have been considered for forecasting the viewership of television channels. An experimental comparison of chosen algorithms showed that the support vector regression algorithm (SVR) gives the best results on the available real world viewership data. The linear regression algorithm does not give satisfactory results, and the extremely randomized trees algorithm (ExtraTrees) has slightly worse results than the support vector machine algorithm. The implementation of a web application that allows users to forecast viewership based on the uploaded broadcast schedule is also described. The application can be used as a schedule optimization tool. Due to the high time complexity of the support vector machine algorithm, the randomized trees algorithm was chosen for use in the application.

• Keywords: strojno učenje; regresija; stroj potpornih vektora; slučajna šuma; web aplikacija
• Keywords (english): machine learning; regression; support vector machine; random forest; web application
• Language: croatian
• URN:NBN: urn:nbn:hr:168:255073
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 12:32:24