Accelerating Infinite Ensemble of Clustering by Pivot Features
The infinite ensemble clustering (IEC) incorporates both ensemble clustering and representation learning by fusing infinite basic partitions and shows appealing performance in the unsupervised context. However, it needs to solve the linear equation system with the high time complexity in proportion to O(d3) where d is the concatenated dimension of many clustering results. Inspired by the cognitive characteristic of human memory that can pay attention to the pivot features in a more compressed data space, we propose an acceleration version of IEC (AIEC) by extracting the pivot features and learning the multiple mappings to reconstruct them, where the linear equation system can be solved with the time complexity O(dr2) (r ≪ d). Experimental results on the standard datasets including image and text ones show that our algorithm AIEC improves the running time of IEC greatly but achieves the comparable clustering performance.
Duke Scholars
Published In
DOI
EISSN
ISSN
Publication Date
Volume
Issue
Start / End Page
Related Subject Headings
- 1702 Cognitive Sciences
- 1109 Neurosciences
- 0801 Artificial Intelligence and Image Processing
Citation
Published In
DOI
EISSN
ISSN
Publication Date
Volume
Issue
Start / End Page
Related Subject Headings
- 1702 Cognitive Sciences
- 1109 Neurosciences
- 0801 Artificial Intelligence and Image Processing