Subspace Determination Through Local Intrinsic Dimensional Decomposition

Ruben Becker, Imane Hafnaoui, Michael E. Houle, Pan Li, Arthur Zimek

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

84 Downloads (Pure)


Axis-aligned subspace clustering generally entails searching through enormous numbers of subspaces (feature combinations) and evaluation of cluster quality within each subspace. In this paper, we tackle the problem of identifying subsets of features with the most significant contribution to the formation of the local neighborhood surrounding a given data point. For each point, the recently-proposed Local Intrinsic Dimension (LID) model is used in identifying the axis directions along which features have the greatest local discriminability, or equivalently, the fewest number of components of LID that capture the local complexity of the data. In this paper, we develop an estimator of LID along axis projections, and provide preliminary evidence that this LID decomposition can indicate axis-aligned data subspaces that support the formation of clusters.
Original languageEnglish
Title of host publicationSimilarity Search and Applications - 12th International Conference, SISAP 2019, Newark, NJ, USA, October 2-4, 2019, Proceedings
EditorsGiuseppe Amato, Claudio Gennaro, Vincent Oria, Miloš Radovanovic
Publication date2019
ISBN (Print)978-3-030-32046-1
ISBN (Electronic)978-3-030-32047-8
Publication statusPublished - 2019
EventInternational Conference on Similarity Search and Applications - New Jersey Institute of Technology (NJIT), Newark, United States
Duration: 2. Oct 20194. Oct 2019
Conference number: 12


ConferenceInternational Conference on Similarity Search and Applications
LocationNew Jersey Institute of Technology (NJIT)
Country/TerritoryUnited States
Internet address
SeriesLecture Notes in Computer Science


  • Estimation
  • Intrinsic dimensionality
  • Subspace


Dive into the research topics of 'Subspace Determination Through Local Intrinsic Dimensional Decomposition'. Together they form a unique fingerprint.

Cite this