본문

서브메뉴

Differentially Private Algorithmic Design: Private Bandits, Counting and Histograms
Differentially Private Algorithmic Design: Private Bandits, Counting and Histograms
Differentially Private Algorithmic Design: Private Bandits, Counting and Histograms

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260202104705
ISBN  
9798286499458
DDC  
310
저자명  
Ou, Tingting.
서명/저자  
Differentially Private Algorithmic Design: Private Bandits, Counting and Histograms
발행사항  
[Sl] : Columbia University, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
141 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-01, Section: B.
주기사항  
Advisor: Cummings, Rachel;Avella Medina, Marco.
학위논문주기  
Thesis (Ph.D.)--Columbia University, 2025.
초록/해제  
요약As concerns around data privacy continue to grow, designing differentially private (DP) algorithms has become a central topic of interest, particularly for statistical and machine learning methods used in practice. This thesis contributes to this line of work by presenting new results in private algorithm design and analysis for multi-armed bandits, counting, and histogram estimation.The thesis begins by showing that the classical Thompson Sampling algorithm with a Gaussian prior for multi-armed bandits is inherently differentially private without any modification. We further propose simple modifications that yield improved privacy guarantees, and analyze how these adjustments affect regret. Next, we study distinct count estimation under differential privacy in the turnstile streaming model. We present the first differentially private algorithms using sublinear space in the stream length T. Our method achieves an error and space complexity of \uD835\uDC42˜(\uD835\uDC47 1/3 ), significantly improving over prior linear-space approaches. Lastly, we address locally private succinct histogram estimation in the federated setting. Building on the sample-and-threshold paradigm which is only centrally private, we develop a locally private algorithm that incorporates an additional local randomization step. This new method achieves stronger privacy guarantees without significantly compromising the accuracy.
일반주제명  
Statistics
일반주제명  
Applied mathematics
일반주제명  
Computer science
키워드  
Histograms
키워드  
Thompson Sampling algorithm
키워드  
Differentially private algorithms
키워드  
Differential privacy
키워드  
Turnstile streaming model
기타저자  
Columbia University Operations Research
기본자료저록  
Dissertations Abstracts International. 87-01B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017358458
■00520260202104705
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798286499458
■035    ▼a(MiAaPQ)AAI32116890
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a310
■1001  ▼aOu,  Tingting.
■24510▼aDifferentially  Private  Algorithmic  Design:  Private  Bandits,  Counting  and  Histograms
■260    ▼a[Sl]▼bColumbia  University▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a141  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-01,  Section:  B.
■500    ▼aAdvisor:  Cummings,  Rachel;Avella  Medina,  Marco.
■5021  ▼aThesis  (Ph.D.)--Columbia  University,  2025.
■520    ▼aAs  concerns  around  data  privacy  continue  to  grow,  designing  differentially  private  (DP)  algorithms  has  become  a  central  topic  of  interest,  particularly  for  statistical  and  machine  learning  methods  used  in  practice.  This  thesis  contributes  to  this  line  of  work  by  presenting  new  results  in  private  algorithm  design  and  analysis  for  multi-armed  bandits,  counting,  and  histogram  estimation.The  thesis  begins  by  showing  that  the  classical  Thompson  Sampling  algorithm  with  a  Gaussian  prior  for  multi-armed  bandits  is  inherently  differentially  private  without  any  modification.  We  further  propose  simple  modifications  that  yield  improved  privacy  guarantees,  and  analyze  how  these  adjustments  affect  regret.  Next,  we  study  distinct  count  estimation  under  differential  privacy  in  the  turnstile  streaming  model.  We  present  the  first  differentially  private  algorithms  using  sublinear  space  in  the  stream  length  T.  Our  method  achieves  an  error  and  space  complexity  of  \uD835\uDC42˜(\uD835\uDC47  1/3  ),  significantly  improving  over  prior  linear-space  approaches.  Lastly,  we  address  locally  private  succinct  histogram  estimation  in  the  federated  setting.  Building  on  the  sample-and-threshold  paradigm  which  is  only  centrally  private,  we  develop  a  locally  private  algorithm  that  incorporates  an  additional  local  randomization  step.  This  new  method  achieves  stronger  privacy  guarantees  without  significantly  compromising  the  accuracy.
■590    ▼aSchool  code:  0054.
■650  4▼aStatistics
■650  4▼aApplied  mathematics
■650  4▼aComputer  science
■653    ▼aHistograms
■653    ▼aThompson  Sampling  algorithm
■653    ▼aDifferentially  private  algorithms
■653    ▼aDifferential  privacy
■653    ▼aTurnstile  streaming  model
■690    ▼a0796
■690    ▼a0984
■690    ▼a0463
■690    ▼a0364
■71020▼aColumbia  University▼bOperations  Research.
■7730  ▼tDissertations  Abstracts  International▼g87-01B.
■790    ▼a0054
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17358458▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

미리보기

내보내기

chatGPT토론

Ai 추천 관련 도서


    신착도서 더보기
    최근 3년간 통계입니다.

    소장정보

    • 예약
    • 소재불명신고
    • 나의폴더
    • 우선정리요청
    • 비도서대출신청
    • 야간 도서대출신청
    소장자료
    등록번호 청구기호 소장처 대출가능여부 대출정보
    TF16035 전자도서 대출가능 마이폴더 부재도서신고 비도서대출신청 야간 도서대출신청

    * 대출중인 자료에 한하여 예약이 가능합니다. 예약을 원하시면 예약버튼을 클릭하십시오.

    해당 도서를 다른 이용자가 함께 대출한 도서

    관련 인기도서

    로그인 후 이용 가능합니다.