TY - GEN
T1 - Robust dirichlet process mixtures
AU - Sun, Jianyong
AU - Garibaldi, Jonathan M.
PY - 2011
Y1 - 2011
N2 - Non-parametric Dirichlet Process mixture (DPM) approaches for density estimation and clustering allow for automatic model selection. In this paper, we aim to develop robust DPM algorithm for clustering datasets with scatter objects, or outliers. In the developed mean-field variational inference algorithms, the auxiliary posterior distributions are factorized in a tree-structured form. In the experiments, we first show the advantage of the tree-structured factorization over the commonly-used full factorization. Then the performances of the robust DPM is evaluated using controlled experiment settings. Finally, the developed robust DPM is applied to biology datasets.
AB - Non-parametric Dirichlet Process mixture (DPM) approaches for density estimation and clustering allow for automatic model selection. In this paper, we aim to develop robust DPM algorithm for clustering datasets with scatter objects, or outliers. In the developed mean-field variational inference algorithms, the auxiliary posterior distributions are factorized in a tree-structured form. In the experiments, we first show the advantage of the tree-structured factorization over the commonly-used full factorization. Then the performances of the robust DPM is evaluated using controlled experiment settings. Finally, the developed robust DPM is applied to biology datasets.
UR - https://www.scopus.com/pages/publications/80053416677
U2 - 10.1109/ICNC.2011.6022352
DO - 10.1109/ICNC.2011.6022352
M3 - 会议稿件
AN - SCOPUS:80053416677
SN - 9781424499533
T3 - Proceedings - 2011 7th International Conference on Natural Computation, ICNC 2011
SP - 1556
EP - 1560
BT - Proceedings - 2011 7th International Conference on Natural Computation, ICNC 2011
T2 - 2011 7th International Conference on Natural Computation, ICNC 2011
Y2 - 26 July 2011 through 28 July 2011
ER -