서브메뉴
검색
Optimizing AI Models for Human Use
Optimizing AI Models for Human Use
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202104852
- ISBN
- 9798288816536
- DDC
- 621.3
- 서명/저자
- Optimizing AI Models for Human Use
- 발행사항
- [Sl] : Stanford University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 151 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-02, Section: B.
- 주기사항
- Advisor: Zou, James.
- 학위논문주기
- Thesis (Ph.D.)--Stanford University, 2025.
- 초록/해제
- 요약AI models are increasingly deployed for human use, from clinical diagnostic tools to general-purpose assistants. The growing prevalence and capability of these models necessitates a deeper understanding of human-AI interaction. In this thesis, I argue that effective human-AI interaction requires integrating human users directly into the design and optimization of AI systems. First, I present my work developing an AI mechanism providing real-time guidance to dermatology patients, demonstrably improving the quality of data collected for telemedicine. Next, I describe my work developing an algorithmic model for human behavior and show how this model can be used to modify the AI's objective function and optimize the model for better joint human-AI performance. I also present my work analyzing user preference and interaction patterns with generative models to understand interaction strategies, quantify AI steerability, and better characterize human users, critical for personalization and model evaluation. Across these studies, I argue that it is imperative to treat humans as an integral part of the model design and optimization process to enable more collaborative human-AI systems.
- 일반주제명
- Computer engineering
- 일반주제명
- Computer science
- 키워드
- Human users
- 키워드
- Model evaluation
- 기타저자
- Stanford University.
- 기본자료저록
- Dissertations Abstracts International. 87-02B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017359230
■00520260202104852
■006m o d
■007cr#unu||||||||
■020 ▼a9798288816536
■035 ▼a(MiAaPQ)AAI32200980
■035 ▼a(MiAaPQ)Stanfordqf711hq4666
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a621.3
■1001 ▼aVodrahalli, Kailas.
■24510▼aOptimizing AI Models for Human Use
■260 ▼a[Sl]▼bStanford University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a151 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-02, Section: B.
■500 ▼aAdvisor: Zou, James.
■5021 ▼aThesis (Ph.D.)--Stanford University, 2025.
■520 ▼aAI models are increasingly deployed for human use, from clinical diagnostic tools to general-purpose assistants. The growing prevalence and capability of these models necessitates a deeper understanding of human-AI interaction. In this thesis, I argue that effective human-AI interaction requires integrating human users directly into the design and optimization of AI systems. First, I present my work developing an AI mechanism providing real-time guidance to dermatology patients, demonstrably improving the quality of data collected for telemedicine. Next, I describe my work developing an algorithmic model for human behavior and show how this model can be used to modify the AI's objective function and optimize the model for better joint human-AI performance. I also present my work analyzing user preference and interaction patterns with generative models to understand interaction strategies, quantify AI steerability, and better characterize human users, critical for personalization and model evaluation. Across these studies, I argue that it is imperative to treat humans as an integral part of the model design and optimization process to enable more collaborative human-AI systems.
■590 ▼aSchool code: 0212.
■650 4▼aComputer engineering
■650 4▼aComputer science
■653 ▼aHuman users
■653 ▼aModel evaluation
■653 ▼aClinical diagnostic tools
■690 ▼a0800
■690 ▼a0984
■690 ▼a0464
■71020▼aStanford University.
■7730 ▼tDissertations Abstracts International▼g87-02B.
■790 ▼a0212
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17359230▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
