Download presentation
Presentation is loading. Please wait.
Published byWilliam Bond Modified over 9 years ago
1
Clustering With Constraints Feasibility Issues and the k-Means Algorithm 報告者:林俞均 日期: 2014/8/27
2
Abstract the CVQE(Constrained Vector Quantization Error) algorithm was proposed in this context It modifies the objective function of traditional K-means CVQE was shown to efficiently produce high- quality clustering of UCI data.
3
Constrained K-means Algorithm Must-link and cannot-link Must-link(ML): constraints specify that two instances have to be in the same cluster Cannot-link(CL): constraints specify that two instances must not be placed in the same cluster
4
The CVQE Algorithm k-means algorithm CVQE algorithm
5
The remaining terms are the errors associated with the must-link (ml=1) and cannot-link (ml=0) constraints. We use ¬ to denote the negation of the Delta function.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.