Music recommendations with deep learning
(2019) EITM01 20172Department of Electrical and Information Technology
- Abstract
- In this thesis we apply deep reinforcement learning to the problem of recom-
mending music. A content-based approach is taken, and features from music
is extracted with a pretrained deep learning music-tagger. For training, user-
interactions are simulated. - Popular Abstract
- In this thesis multiple machine learning, and specifically deep learning techniques were used to recommend music to users interacting with a music recommendation system. To recommend music a user enjoys as soon as possible deep reinforcement learning was applied, and because no real users were available we trained the system with simulated users. For reducing the amount of information in music, features were extracted from a pre-trained music-tagger.
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/8993717
- author
- Rynegardh, Carl LU
- supervisor
- organization
- course
- EITM01 20172
- year
- 2019
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Reinforcement learning, deep learning, music information retrieval
- report number
- LU/LTH-EIT 2019-718
- language
- English
- id
- 8993717
- date added to LUP
- 2019-09-10 09:35:34
- date last changed
- 2019-09-10 09:35:34
@misc{8993717, abstract = {{In this thesis we apply deep reinforcement learning to the problem of recom- mending music. A content-based approach is taken, and features from music is extracted with a pretrained deep learning music-tagger. For training, user- interactions are simulated.}}, author = {{Rynegardh, Carl}}, language = {{eng}}, note = {{Student Paper}}, title = {{Music recommendations with deep learning}}, year = {{2019}}, }