Click Here to Join via Zoom
Meeting # 942 7359 5552

Zoom Meeting # 942 7359 5552

Abstract:

Subspace outlier detection is a fundamental data mining task for high-dimensional data. The existing hashing-based methods only effectively identify outliers in axis-parallel subspaces but fail on arbitrary-shaped, schema-less data such as time series and graphs. We present OS-Hash (Oblique Subspace Hashing), a linear-time, constant-space solution that defines oblique subspaces based on pairwise object proximity, without requiring explicit multidimensional representations. By hashing data into arbitrarily oriented subspaces, OS-Hash builds efficient histograms for outlier detection across diverse data types and extends naturally to the data stream setting.

Speaker Biography:

Bowen Li is a final-year Ph.D. student in Computer Science at Florida State University, advised by Professor Peixiang Zhao. Before pursuing my doctoral degree, he earned a B.E. in Computer Science from East China University of Science and Technology and an M.S. in Computer Science from Florida State University. His research interests span data and network science, database systems, and core data mining problems in machine learning. He also gained industry experience as a summer intern with the Relational Database Service (RDS) team at Amazon Web Services (AWS) in 2022 and 2023.