Institutional Repository of Vilnius University
Reinforcement learning for hard dynamic programming problems /
Abstract
dc:descriptionThis thesis investigates using Reinforcement Learning instead of calculating minimum energy seams in dynamic programming algorithm seam carving. After explaining selected methodology experiments are conducted to find viable Reinforcement Learning environment and train agent, which would be able to find minimal energy sum vertical seam in the image. Analysis of the results show that while Reinforcement Learning variation of seam carving performs worse due to overfitting and has long initial training time, its execution time is much faster and scales much better than the original algorithm.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Čiplys, Arnoldas,
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD146236061&prefLang=en_US