แจ้งเอกสารไม่ครบถ้วน, ไม่ตรงกับชื่อเรื่อง หรือมีข้อผิดพลาดเกี่ยวกับเอกสาร ติดต่อที่นี่ ==>
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
หากไม่มีอีเมลผู้รับให้กรอก thailis-noc@uni.net.th
Seksan Kiatsupaibul. Algorithm comparison between Thompson sampling and upper confidence bound for sequential decision problem in the game of Rock-Paper-Scissors. (). King Mongkut's University of Technology North Bangkok. Central Library. : , 2022.
Algorithm comparison between Thompson sampling and upper confidence bound for sequential decision problem in the game of Rock-Paper-Scissors
Name: Thanyavuth Akarasomcheep
Organization :
Chulalongkorn University. Faculty of Commerce and Accountancy
Email :
thanyavuth@outlook.com
Name: Seksan Kiatsupaibul
Organization :
Chulalongkorn University. Faculty of Commerce and Accountancy
Email :
seksan@cbs.chula.ac.th
keyword:
Thompson Sampling
ThaSH:
Reinforcement learning
ThaSH:
Markov processes
ThaSH:
Decision support systems
Abstract:
The Thompson sampling algorithm and the upper confidence bound algorithm are known to be two efficient
algorithms for solving multi-armed bandit problems and some Markov decision processes. However, it is unclear
how well those two algorithms perform when encountering human decisions that involve behavioral components.
This study set up a simple human decision problem in the form of a series of simulated rock-paper-scissors games.
Two human behavioral traits, a mixed clockwise strategy and a stop loss strategy, are simulated in the games. Two
reinforcement learning agents, one with the Thompson sampling algorithm and another one with the upper
confidence bound algorithm, are designed by modeling the behavioral problem as a Markov decision process. Under
different parameter settings, the performances of the two agents solving the problem are compared, taking the
cumulative reward as the performance measure. We find that the upper confidence bound agent outperforms the
Thompson sampling agent in most of the experimental settings. The exception where the Thompson sampling
outperforms the upper confidence bound is when there exists a strong behavioral pattern in the case of a long decision horizon.
King Mongkut's University of Technology North Bangkok. Central Library
Address:
BANGKOK
Email:
library@kmutnb.ac.th
Created:
2022
Modified:
2025-09-17
Issued:
2025-09-17
บทความ/Article
application/pdf
BibliograpyCitation :
In King Mongkut's University of Technology North Bangkok Faculty of Applied Science, Thai Statistical Association (TSA) and Statistics Cooperative Research Network (Statistics CRN). The Proceeding of International Conference on Applied Statistics (ICAS 2022) (pp.208-213). Bangkok : King Mongkut's University of Technology North Bangkok
eng
©copyrights King Mongkut's University of Technology North Bangkok
RightsAccess:
ใช้เวลา
0.032234 วินาที
Thanyavuth Akarasomcheep
| Title | Contributor | Type |
|---|---|---|
| Algorithm comparison between Thompson sampling and upper confidence bound for sequential decision problem in the game of Rock-Paper-Scissors
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Thanyavuth Akarasomcheep;Seksan Kiatsupaibul | บทความ/Article |
Seksan Kiatsupaibul
| Title | Contributor | Type |
|---|---|---|
| An application of reinforcement learning to credit scoring based on the Logistic Bandit framework
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Kantapong Visantavarakul;Seksan Kiatsupaibul | บทความ/Article | |
| An efficiency comparison of distance measures in K-Nearest Neighbors imputation for spatial data
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Prawit Banjong;Seksan Kiatsupaibul | บทความ/Article | |
| Algorithm comparison between Thompson sampling and upper confidence bound for sequential decision problem in the game of Rock-Paper-Scissors
มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าพระนครเหนือ Thanyavuth Akarasomcheep;Seksan Kiatsupaibul | บทความ/Article |