Adaptive Self-Optimal Softmax Clustering

MPhil Thesis Defence


Title: "Adaptive Self-Optimal Softmax Clustering"

By

Mr. Ziyao ZHANG


Abstract

Discriminative clustering approaches assign data points into different groups 
by identifying sparse regions, without modeling the dataset and categories 
explicitly. Such methods are flexible and powerful in practice since they make 
few assumptions. In particular, the probabilistic-based Softmax model makes 
only one assumption that data points are linearly separable, so it is 
potentially suitable in clustering data preprocessed by feature transformation 
techniques. The principle of cluster assumption states that decision boundaries 
of clusters should lie in low-density regions. In previous works on 
discriminative clustering, this principle is compromised by the cluster balance 
consideration, which is incorporated to avoid degenerate clustering solutions. 
However, datasets are rarely balanced with respect to attributes of interest. 
Furthermore, large clusters from imbalanced datasets might also contain sparse 
regions, where decision boundaries should not be positioned. In this thesis, we 
present self-optimality, a novel criterion for Softmax discriminative 
clustering that is faithful to the cluster assumption principle and is free of 
the cluster balance consideration. We also propose an adaptive algorithm aiming 
at finding self-optimal solutions, which can accurately recognize clusters from 
imbalanced datasets with multiple degrees of sparseness.


Date:  			Wednesday, 16 December 2020

Time:			2:00pm - 4:00pm

Zoom meeting:		https://hkust.zoom.us/j/6761083097

Committee Members:	Prof. Nevin Zhang (Supervisor)
 			Prof. Raymond Wong (Chairperson)
 			Dr. Dan Xu


**** ALL are Welcome ****