가야대학교 분성도서관

상단 글로벌/추가 메뉴

회원 로그인

주메뉴

: • 전체검색; • 단행본; • 정기간행물; • 학위논문; • 비도서; • 특정번호 검색; • 기사색인; • 학습연구지원; • 지정도서

: • 전자자원통합검색; • 해외 WEB DB; • E-Book; • E-Learning; • 국내 WEB DB; • 인터넷검색

: • 대출현황조회/연장; • 예약현황조회/취소; • 희망도서; • 내서재; • 나의서평; • 나의태그; • 나의위젯; • RSS; • SDI; • 개인정보관리

: • 공지사항; • FAQ; • Q&A 일반·학술; • 신착/인기 도서; • 학위논문온라인제출; • 외부인 이용안내; • 설문조사

: • 도서관 소개; • 연혁; • 현황; • 이용안내; • 찾아오시는 길

자료검색

Home
상세정보

상세정보

검색결과 돌아가기

부가기능

MARC보기

Characterizing the Limits and Defenses of Machine Learning in Adversarial Settings

상세 프로파일

상세정보
자료유형	E-Book
개인저자	Papernot, Nicolas.
단체저자명	The Pennsylvania State University. Computer Science and Engineering.
서명/저자사항	Characterizing the Limits and Defenses of Machine Learning in Adversarial Settings.
발행사항	[S.l.] : The Pennsylvania State University., 2018
발행사항	Ann Arbor : ProQuest Dissertations & Theses, 2018
형태사항	178 p.
소장본 주기	School code: 0176.
ISBN	9780438135536
일반주기	Source: Dissertation Abstracts International, Volume: 79-12(E), Section: B.
요약	Advances in machine learning (ML) in recent years have enabled a dizzying array of applications such as object recognition, autonomous systems, security diagnostics, and playing the game of Go. Machine learning is not only a new paradigm for bui
요약	In this thesis, I focus my study on the integrity of ML models. Integrity refers here to the faithfulness of model predictions with respect to an expected outcome. This property is at the core of traditional machine learning evaluation, as demon
요약	A large fraction of ML techniques were designed for benign execution environments. Yet, the presence of adversaries may invalidate some of these underlying assumptions by forcing a mismatch between the distributions on which the model is trained
요약	I explore the space of attacks against ML integrity at test time. Given full or limited access to a trained model, I devise strategies that modify the test data to create a worst-case drift between the training and test distributions. The implic
요약	Hence, my efforts to increase the robustness of models to these adversarial manipulations strive to decrease the confidence of predictions made far from the training distribution. Informed by my progress on attacks operating in the black-box thr
요약	I then describe recent defensive efforts addressing these shortcomings. To this end, I introduce the Deep k-Nearest Neighbors classifier, which augments deep neural networks with an integrity check at test time. The approach compares internal re
요약	This research calls for future efforts to investigate the robustness of individual layers of deep neural networks rather than treating the model as a black-box. This aligns well with the modular nature of deep neural networks, which orchestrate
일반주제명	Computer science.
언어	영어
기본자료 저록	Dissertation Abstracts International79-12B(E). Dissertation Abstract International
대출바로가기	http://www.riss.kr/pdu/ddodLink.do?id=T15000703