M-Zoom: Fast Dense-Block Detection in Tensors with Quality Guarantees

Overview

M-Zoom (Multidimensional Zoom) and M-Biz (Multidimensional Bi-directional Zoom) are algorithms for detecting dense subtensors.
They have the following properties:

• Scalable: scales almost linearly with all input factors
• Provably accurate: provides high accuracy in real data as well as theoretical guarantees
• Flexible: supports high-order tensors, various density measures, multi-subtensor detection, and size bounds

Papers

M-Zoom and M-Biz are described in the following papers:

• M-Zoom: Fast Dense-Block Detection in Tensors with Quality Guarantees.
  Kijung Shin, Bryan Hooi, and Christos Faloutsos.
  European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML-PKDD) 2016, Riva del Garda, Italy
  [PDF] [Supplementary Document] [BIBTEX]


• Fast, Accurate and Flexible Algorithms for Dense Subtensor Mining.
  Kijung Shin, Bryan Hooi, and Christos Faloutsos.
  ACM Transactions on Knowledge Discovery from Data (TKDD), vol. 12, no. 3, pp. 28:1-28:30, January 2018
  [PDF] [BIBTEX]

Codes

The source code used in the papers is available. [GitHub Repository]

Datasets

People

• Kijung Shin (Carnegie Mellon University)
• Bryan Hooi (Carnegie Mellon University)
• Christos Faloutsos (Carnegie Mellon University)