M
Michael E. Houle
Researcher at National Institute of Informatics
Publications - 124
Citations - 4574
Michael E. Houle is an academic researcher from National Institute of Informatics. The author has contributed to research in topics: Curse of dimensionality & Nearest neighbor search. The author has an hindex of 31, co-authored 122 publications receiving 3867 citations. Previous affiliations of Michael E. Houle include University of Sydney & McGill University.
Papers
More filters
Journal ArticleDOI
On the evaluation of unsupervised outlier detection: measures, datasets, and an empirical study
Guilherme Oliveira Campos,Arthur Zimek,Jörg Sander,Ricardo J. G. B. Campello,Barbora Micenková,Erich Schubert,Ira Assent,Michael E. Houle +7 more
TL;DR: An extensive experimental study on the performance of a representative set of standard k nearest neighborhood-based methods for unsupervised outlier detection, across a wide variety of datasets prepared for this purpose, and provides a characterization of the datasets themselves.
Posted Content
Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality
Xingjun Ma,Bo Li,Yisen Wang,Sarah M. Erfani,Sudanthi Wijewickrema,Grant Schoenebeck,Dawn Song,Michael E. Houle,James Bailey +8 more
TL;DR: The analysis of the LID characteristic for adversarial regions not only motivates new directions of effective adversarial defense, but also opens up more challenges for developing new attacks to better understand the vulnerabilities of DNNs.
Proceedings Article
Characterizing Adversarial Subspaces Using Local Intrinsic Dimensionality
Xingjun Ma,Bo Li,Yisen Wang,Sarah M. Erfani,Sudanthi Wijewickrema,Grant Schoenebeck,Dawn Song,Michael E. Houle,James Bailey +8 more
TL;DR: In this article, the dimensional properties of adversarial regions are characterized via the use of Local Intrinsic Dimensionality (LID), which assesses the space-filling capability of the region surrounding a reference example, based on the distance distribution of the example to its neighbors.
Book ChapterDOI
Can shared-neighbor distances defeat the curse of dimensionality?
TL;DR: It is found that rank-based similarity measures can result in more stable performance than their associated primary distance measures.
Patent
Computer system, method, and program product for generating a data structure for information retrieval, and an associated graphical user interface
TL;DR: In this article, a computer system for generating data structures for information retrieval of documents stored in a database is described, which includes: a neighborhood patch generation system for defining patch of nodes having predetermined similarities in a hierarchy structure.