본문

서브메뉴

Conditional Guarantees in Model-Free Inference
Conditional Guarantees in Model-Free Inference
Conditional Guarantees in Model-Free Inference

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260202104740
ISBN  
9798290651453
DDC  
305.8
저자명  
Cherian, John.
서명/저자  
Conditional Guarantees in Model-Free Inference
발행사항  
[Sl] : Stanford University, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
180 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-03, Section: B.
주기사항  
Advisor: Candès, Emmanuel.
학위논문주기  
Thesis (Ph.D.)--Stanford University, 2025.
초록/해제  
요약The deployment of machine learning in high-stakes settings has raised fundamental questions about the reliability and fairness of black-box models. For example, does a model treat different groups equitably, and can we quantify model uncertainty before taking action on each prediction? While numerous assumption-lean methods appear to address these questions, their guarantees often fall short in practice. For example, model-free guarantees in predictive inference are often marginal, meaning that the assessment of prediction error holds only on-average for a data point drawn exchangeably from the data used to calibrate the method. These guarantees, therefore, do not rule out the existence of substantial individual variation. The research program presented in this dissertation aims at the inherent tension of model-free statistical inference: the generic validity of such methods is appealing, but without a well-specified model, it is challenging to identify guarantees that are also practically useful.1.1 Fairness auditingWhile black-box models may demonstrate impressive accuracy on average, their performance can still vary substantially between subpopulations. For example, an algorithm deployed for recidivism prediction exhibits significantly higher false positive rates for African-American relative to Caucasian parolees [Angwin et al., 2016]. Similar performance disparities have been documented in other highstakes applications such as facial recognition and hiring [Buolamwini and Gebru, 2018, Dastin, 2018].Motivated by this concern, numerous stakeholders have solicited methods, often referred to as "fairness audits," that can discover and quantify such disparities [Brundage et al., 2020, Schaake and Clark, 2022]. Despite substantial prior work in this area [DiCiccio et al., 2020, Morina et al., 2019, Si et al., 2021, Taskesen et al., 2021, Tramer et al., 2017, von Zahn et al., 2023, Xue et al., 2020, Yan and Zhang, 2022], the definition of fairness auditing remains fraught. Fairness auditing is often framed as a single statistical test that rejects in the case of any performance disparity over a limited set of sensitive subpopulations [DiCiccio et al., 2020, Morina et al., 2019, Roy and Mohapatra, 2023, Si et al., 2021, Taskesen et al., 2021, Tramer et al., 2017, Xue et al., 2020]. While follow-up investigation to localize disparities is desired (and often performed), this task raises new challenges. For example, empirical parity across a limited collection of subgroups does not rule out substantial disparities among smaller subgroups [Kearns et al., 2018]. Further, if we consider multiple performance metrics over a rich collection of subpopulations, discovering some disparity between two subgroups is hardly surprising. Unfortunately, existing methods for identifying localized (dis-)parities are accompanied by few statistical guarantees [Schaake and Clark, 2022, von Zahn et al., 2023, Yan and Zhang, 2022].In Chapter 2, we develop a family of statistical methods that rigorously achieve two goals: (1) the "certification" of subpopulations for which the model performs adequately, and (2) the "flagging" of subpopulations that suffer harmful performance disparities. Formally, we approach these two tasks by allowing the auditor to define a "disparity" by comparing some measure of model performance on a subpopulation to a potentially data-dependent target. A certification audit then allows the auditor to identify subpopulations for which this disparity is acceptably low, while a flagging audit discovers subpopulations for which this disparity exceeds some prespecified threshold. Our proposed methods only require access to a so-called "audit trail," i.e., model predictions on a data set held out from training [Brundage et al., 2020], but not white-box access to the model itself.
일반주제명  
White people
일반주제명  
Statistical inference
일반주제명  
Conformity
일반주제명  
Large language models
일반주제명  
Hilbert space
일반주제명  
Certification
일반주제명  
Computer engineering
키워드  
Black-box models
키워드  
Machine learning
기타저자  
Stanford University.
기본자료저록  
Dissertations Abstracts International. 87-03B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017358705
■00520260202104740
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798290651453
■035    ▼a(MiAaPQ)AAI32149693
■035    ▼a(MiAaPQ)Stanfordmx648rd8955
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a305.8
■1001  ▼aCherian,  John.
■24510▼aConditional  Guarantees  in  Model-Free  Inference
■260    ▼a[Sl]▼bStanford  University▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a180  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-03,  Section:  B.
■500    ▼aAdvisor:  Candès,  Emmanuel.
■5021  ▼aThesis  (Ph.D.)--Stanford  University,  2025.
■520    ▼aThe  deployment  of  machine  learning  in  high-stakes  settings  has  raised  fundamental  questions  about  the  reliability  and  fairness  of  black-box  models.  For  example,  does  a  model  treat  different  groups  equitably,  and  can  we  quantify  model  uncertainty  before  taking  action  on  each  prediction?  While  numerous  assumption-lean  methods  appear  to  address  these  questions,  their  guarantees  often  fall  short  in  practice.  For  example,  model-free  guarantees  in  predictive  inference  are  often  marginal,  meaning  that  the  assessment  of  prediction  error  holds  only  on-average  for  a  data  point  drawn  exchangeably  from  the  data  used  to  calibrate  the  method.  These  guarantees,  therefore,  do  not  rule  out  the  existence  of  substantial  individual  variation.  The  research  program  presented  in  this  dissertation  aims  at  the  inherent  tension  of  model-free  statistical  inference:  the  generic  validity  of  such  methods  is  appealing,  but  without  a  well-specified  model,  it  is  challenging  to  identify  guarantees  that  are  also  practically  useful.1.1  Fairness  auditingWhile  black-box  models  may  demonstrate  impressive  accuracy  on  average,  their  performance  can  still  vary  substantially  between  subpopulations.  For  example,  an  algorithm  deployed  for  recidivism  prediction  exhibits  significantly  higher  false  positive  rates  for  African-American  relative  to  Caucasian  parolees  [Angwin  et  al.,  2016].  Similar  performance  disparities  have  been  documented  in  other  highstakes  applications  such  as  facial  recognition  and  hiring  [Buolamwini  and  Gebru,  2018,  Dastin,  2018].Motivated  by  this  concern,  numerous  stakeholders  have  solicited  methods,  often  referred  to  as  "fairness  audits,"  that  can  discover  and  quantify  such  disparities  [Brundage  et  al.,  2020,  Schaake  and  Clark,  2022].  Despite  substantial  prior  work  in  this  area  [DiCiccio  et  al.,  2020,  Morina  et  al.,  2019,  Si  et  al.,  2021,  Taskesen  et  al.,  2021,  Tramer  et  al.,  2017,  von  Zahn  et  al.,  2023,  Xue  et  al.,  2020,  Yan  and  Zhang,  2022],  the  definition  of  fairness  auditing  remains  fraught.  Fairness  auditing  is  often  framed  as  a  single  statistical  test  that  rejects  in  the  case  of  any  performance  disparity  over  a  limited  set  of  sensitive  subpopulations  [DiCiccio  et  al.,  2020,  Morina  et  al.,  2019,  Roy  and  Mohapatra,  2023,  Si  et  al.,  2021,  Taskesen  et  al.,  2021,  Tramer  et  al.,  2017,  Xue  et  al.,  2020].  While  follow-up  investigation  to  localize  disparities  is  desired  (and  often  performed),  this  task  raises  new  challenges.  For  example,  empirical  parity  across  a  limited  collection  of  subgroups  does  not  rule  out  substantial  disparities  among  smaller  subgroups  [Kearns  et  al.,  2018].  Further,  if  we  consider  multiple  performance  metrics  over  a  rich  collection  of  subpopulations,  discovering  some  disparity  between  two  subgroups  is  hardly  surprising.  Unfortunately,  existing  methods  for  identifying  localized  (dis-)parities  are  accompanied  by  few  statistical  guarantees  [Schaake  and  Clark,  2022,  von  Zahn  et  al.,  2023,  Yan  and  Zhang,  2022].In  Chapter  2,  we  develop  a  family  of  statistical  methods  that  rigorously  achieve  two  goals:  (1)  the  "certification"  of  subpopulations  for  which  the  model  performs  adequately,  and  (2)  the  "flagging"  of  subpopulations  that  suffer  harmful  performance  disparities.  Formally,  we  approach  these  two  tasks  by  allowing  the  auditor  to  define  a  "disparity"  by  comparing  some  measure  of  model  performance  on  a  subpopulation  to  a  potentially  data-dependent  target.  A  certification  audit  then  allows  the  auditor  to  identify  subpopulations  for  which  this  disparity  is  acceptably  low,  while  a  flagging  audit  discovers  subpopulations  for  which  this  disparity  exceeds  some  prespecified  threshold.  Our  proposed  methods  only  require  access  to  a  so-called  "audit  trail,"  i.e.,  model  predictions  on  a  data  set  held  out  from  training  [Brundage  et  al.,  2020],  but  not  white-box  access  to  the  model  itself.
■590    ▼aSchool  code:  0212.
■650  4▼aWhite  people
■650  4▼aStatistical  inference
■650  4▼aConformity
■650  4▼aLarge  language  models
■650  4▼aHilbert  space
■650  4▼aCertification
■650  4▼aComputer  engineering
■653    ▼aBlack-box  models
■653    ▼aMachine  learning
■690    ▼a0464
■690    ▼a0800
■71020▼aStanford  University.
■7730  ▼tDissertations  Abstracts  International▼g87-03B.
■790    ▼a0212
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17358705▼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
    TF19249 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.