Towards Novelty-Resilient AI: Learning in the Open World
Towards Novelty-Resilient AI: Learning in the Open World
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211152048
- ISBN
- 9798342104876
- DDC
- 006
- 저자명
- Bonjour, Trevor.
- 서명/저자
- Towards Novelty-Resilient AI: Learning in the Open World
- 발행사항
- [Sl] : Purdue University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 102 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-04, Section: B.
- 주기사항
- Advisor: Bhargava, Bharat.
- 학위논문주기
- Thesis (Ph.D.)--Purdue University, 2024.
- 초록/해제
- 요약Current artificial intelligence (AI) systems are proficient at tasks in a closed-world setting where the rules are often rigid. However, in real-world applications, the environment is usually open and dynamic. In this work, we investigate the effects of such dynamic environments on AI systems and develop ways to mitigate those effects. Central to our exploration is the concept of \extit{novelties}. Novelties encompass structural changes, unanticipated events, and environmental shifts that can confound traditional AI systems. We categorize novelties based on their representation, anticipation, and impact on agents, laying the groundwork for systematic detection and adaptation strategies. We explore novelties in the context of stochastic games. Decision-making in stochastic games exercises many aspects of the same reasoning capabilities needed by AI agents acting in the real world. A multi-agent stochastic game allows for infinitely many ways to introduce novelty. We propose an extension of the deep reinforcement learning (DRL) paradigm to develop agents that can detect and adapt to novelties in these environments. To address the sample efficiency challenge in DRL, we introduce a hybrid approach that combines fixed-policy methods with traditional DRL techniques, offering enhanced performance in complex decision-making tasks. We present a novel method for detecting anticipated novelties in multi-agent games, leveraging information theory to discern patterns indicative of collusion among players. Finally, we introduce DABLER, a pioneering deep reinforcement learning architecture that dynamically adapts to changing environmental conditions through broad learning approaches and environment recognition. Our findings underscore the importance of developing AI systems equipped to navigate the uncertainties of the open world, offering promising pathways for advancing AI research and application in real-world settings.
- 일반주제명
- Deep learning
- 일반주제명
- Pandemics
- 일반주제명
- Decision making
- 일반주제명
- Adaptation
- 일반주제명
- Coronaviruses
- 일반주제명
- Games
- 일반주제명
- Markov analysis
- 일반주제명
- COVID-19
- 일반주제명
- Epidemiology
- 일반주제명
- Virology
- 기타저자
- Purdue University.
- 기본자료저록
- Dissertations Abstracts International. 86-04B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017162742
■00520250211152048
■006m o d
■007cr#unu||||||||
■020 ▼a9798342104876
■035 ▼a(MiAaPQ)AAI31345239
■035 ▼a(MiAaPQ)Purdue25668642
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a006
■1001 ▼aBonjour, Trevor.
■24510▼aTowards Novelty-Resilient AI: Learning in the Open World
■260 ▼a[Sl]▼bPurdue University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a102 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-04, Section: B.
■500 ▼aAdvisor: Bhargava, Bharat.
■5021 ▼aThesis (Ph.D.)--Purdue University, 2024.
■520 ▼aCurrent artificial intelligence (AI) systems are proficient at tasks in a closed-world setting where the rules are often rigid. However, in real-world applications, the environment is usually open and dynamic. In this work, we investigate the effects of such dynamic environments on AI systems and develop ways to mitigate those effects. Central to our exploration is the concept of \extit{novelties}. Novelties encompass structural changes, unanticipated events, and environmental shifts that can confound traditional AI systems. We categorize novelties based on their representation, anticipation, and impact on agents, laying the groundwork for systematic detection and adaptation strategies. We explore novelties in the context of stochastic games. Decision-making in stochastic games exercises many aspects of the same reasoning capabilities needed by AI agents acting in the real world. A multi-agent stochastic game allows for infinitely many ways to introduce novelty. We propose an extension of the deep reinforcement learning (DRL) paradigm to develop agents that can detect and adapt to novelties in these environments. To address the sample efficiency challenge in DRL, we introduce a hybrid approach that combines fixed-policy methods with traditional DRL techniques, offering enhanced performance in complex decision-making tasks. We present a novel method for detecting anticipated novelties in multi-agent games, leveraging information theory to discern patterns indicative of collusion among players. Finally, we introduce DABLER, a pioneering deep reinforcement learning architecture that dynamically adapts to changing environmental conditions through broad learning approaches and environment recognition. Our findings underscore the importance of developing AI systems equipped to navigate the uncertainties of the open world, offering promising pathways for advancing AI research and application in real-world settings.
■590 ▼aSchool code: 0183.
■650 4▼aDeep learning
■650 4▼aPandemics
■650 4▼aDecision making
■650 4▼aAdaptation
■650 4▼aCoronaviruses
■650 4▼aGames
■650 4▼aMarkov analysis
■650 4▼aCOVID-19
■650 4▼aEpidemiology
■650 4▼aVirology
■690 ▼a0800
■690 ▼a0766
■690 ▼a0796
■690 ▼a0720
■71020▼aPurdue University.
■7730 ▼tDissertations Abstracts International▼g86-04B.
■790 ▼a0183
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17162742▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


