Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Predicting Hit Songs with Machine Learning
KTH, School of Electrical Engineering and Computer Science (EECS).
KTH, School of Electrical Engineering and Computer Science (EECS).
2018 (English)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE creditsStudent thesisAlternative title
Förutsäga hitlåtar med maskininlärning (Swedish)
Abstract [en]

Exploring the possibility of predicting hit songs is both interesting from a scientific point of view and something that could be beneficial to the music industry. In this research we raise the question if it is possible to classify a music track as a hit or a non-hit based on its audio features. We investigated which machine learning algorithms could be suited for a task like this. Four different models were built using various algorithms such as Support Vector Machine and Gaussian Naive Bayes. The obtained results do not indicate that it is possible to predict hit songs on our particular dataset. This stands in contrast to some previous research within this field. We discuss the potential problem in using only audio features, and how this seems not to be sufficient information for predicting a hit.

Abstract [sv]

Att förutsäga om en låt når topplistan skulle vara gynnsamt för musikindustrin samtidigt som det är intressant ur ett vetenskapligt perspektiv. I vår undersökning lyfter vi frågan om det är möjligt att klassificera en låt som hit eller icke-hit baserat på låtens ljudegenskaper. Vi utreder vilka algoritmer för maskininlärning som kan vara lämpliga för denna uppgift. Fyra olika modeller byggs utifrån algoritmer såsom Support Vector Machine och Gaussian Naive Bayes. De uppmätta resultaten indikerar att det inte är möjligt att förutsäga en hit på den utvalda datamängden. Dessa resultat motsäger en del av den tidigare forskningen gjord inom detta ämne. Vi diskuterar det potentiella problemet i att bara att analysera ljudegenskaper och hur detta inte tycks vara tillräcklig information för att identifiera en hit.

Place, publisher, year, edition, pages
2018.
Series
TRITA-EECS-EX ; 2018:202
National Category
Computer and Information Sciences
Identifiers
URN: urn:nbn:se:kth:diva-229705OAI: oai:DiVA.org:kth-229705DiVA, id: diva2:1214146
Subject / course
Computer Science
Educational program
Bachelor of Science in Engineering - Computer Engineering
Supervisors
Examiners
Available from: 2018-08-03 Created: 2018-06-05 Last updated: 2018-08-03Bibliographically approved

Open Access in DiVA

fulltext(1055 kB)17 downloads
File information
File name FULLTEXT01.pdfFile size 1055 kBChecksum SHA-512
a5022cd3389933caf048cbb5bf9482808087d13d3de126e6f995afcdddf8e9390f37fcfdddda4c995db70f097a88090d9d41d0185098c9dc998e98f4ac11419d
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Computer and Information Sciences

Search outside of DiVA

GoogleGoogle Scholar
Total: 17 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 25 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf