DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Kim, WooChang | - |
dc.contributor.advisor | 김우창 | - |
dc.contributor.author | Lim, Hyojin | - |
dc.date.accessioned | 2022-04-21T19:31:10Z | - |
dc.date.available | 2022-04-21T19:31:10Z | - |
dc.date.issued | 2021 | - |
dc.identifier.uri | http://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=948509&flag=dissertation | en_US |
dc.identifier.uri | http://hdl.handle.net/10203/295304 | - |
dc.description | 학위논문(석사) - 한국과학기술원 : 산업및시스템공학과, 2021.2,[iii, 32 p. :] | - |
dc.description.abstract | Hedge funds are getting attention as an alternative asset with their attractive performances including absolute return. However, some obstacles still exist regarding liquidity, accessibility and drawdown risk. In this article, we present an advanced framework for index tracking factor model based on integrated factor set and dynamic regularized regression. Rule-based parameter tuning is suggested and tracking error decreased meaningfully. Next, we bring reinforcement learning to factor investment in order to manage drawdown risk of hedge fund index. Deep Q-Learning is utilized to manage maximum drawdown of factor model along with new and delicate definition of reward. In addition, we suggest a modified algorithm by inserting “select-to-update” stage, which makes agent to converge more efficiently. Our DQN based factor model definitely enhanced the return and risk-adjusted return of fund. | - |
dc.language | eng | - |
dc.publisher | 한국과학기술원 | - |
dc.subject | Hedge fund▼aFactor model▼aIndex tracking▼aReinforcement learning▼aDQN▼aMaximum drawdown | - |
dc.subject | 헤지펀드▼a팩터 모델▼a인덱스 추적▼a강화 학습▼aDQN▼a최대 드로우다운 | - |
dc.title | Enhancing hedge fund index tracking factor model via deep reinforcement learning | - |
dc.title.alternative | 심층 강화학습을 통한 헤지 펀드 인덱스 추적 팩터 모델 개선 | - |
dc.type | Thesis(Master) | - |
dc.identifier.CNRN | 325007 | - |
dc.description.department | 한국과학기술원 :산업및시스템공학과, | - |
dc.contributor.alternativeauthor | 임효진 | - |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.