Coresets for Archetypal Analysis
Research output: Contributions to collected editions/works › Article in conference proceedings › Research › peer-review
Standard
32rd Conference on Neural Information Processing Systems (NeurIPS 2019): Vancouver, Canada, 8-14 December 2019. ed. / Hanna Wallach; Hugo Larochelle. Vol. 10 Red Hook: Curran Associates, 2020. p. 7215-7223 (Advances in neural information processing systems; Vol. 32).
Research output: Contributions to collected editions/works › Article in conference proceedings › Research › peer-review
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - CHAP
T1 - Coresets for Archetypal Analysis
AU - Mair, Sebastian
AU - Brefeld, Ulf
N1 - Conference code: 33
PY - 2020
Y1 - 2020
N2 - Archetypal analysis represents instances as linear mixtures of prototypes (the archetypes) that lie on the boundary of the convex hull of the data. Archetypes are thus often better interpretable than factors computed by other matrix factorization techniques. However, the interpretability comes with high computational cost due to additional convexity-preserving constraints. In this paper, we propose efficient coresets for archetypal analysis. Theoretical guarantees are derived by showing that quantization errors of k-means upper bound archetypal analysis; the computation of a provable absolute-coreset can be performed in only two passes over the data. Empirically, we show that the coresets lead to improved performance on several data sets.
AB - Archetypal analysis represents instances as linear mixtures of prototypes (the archetypes) that lie on the boundary of the convex hull of the data. Archetypes are thus often better interpretable than factors computed by other matrix factorization techniques. However, the interpretability comes with high computational cost due to additional convexity-preserving constraints. In this paper, we propose efficient coresets for archetypal analysis. Theoretical guarantees are derived by showing that quantization errors of k-means upper bound archetypal analysis; the computation of a provable absolute-coreset can be performed in only two passes over the data. Empirically, we show that the coresets lead to improved performance on several data sets.
KW - Business informatics
UR - https://papers.nips.cc/paper/8945-coresets-for-archetypal-analysis
UR - https://proceedings.neurips.cc/paper/2019
UR - http://toc.proceedings.com/53719webtoc.pdf
UR - http://www.proceedings.com/53719.html
M3 - Article in conference proceedings
SN - 978-1-71380-793-3
VL - 10
T3 - Advances in neural information processing systems
SP - 7215
EP - 7223
BT - 32rd Conference on Neural Information Processing Systems (NeurIPS 2019)
A2 - Wallach, Hanna
A2 - Larochelle, Hugo
PB - Curran Associates
CY - Red Hook
T2 - 33rd Conference on Neural Information Processing Systems - NeurIPS 2019
Y2 - 8 December 2019 through 14 December 2019
ER -