K-means algorithm based on quasi ideal point method
- Resource Type
- Conference
- Authors
- Liu, Hui-Ming; Bai, Jie; Gan, Chen-Ming
- Source
- 2017 Chinese Automation Congress (CAC) Chinese Automation Congress (CAC), 2017. :2900-2903 Oct, 2017
- Subject
- Aerospace
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Transportation
Clustering algorithms
Partitioning algorithms
Iris
Dispersion
Algorithm design and analysis
Linear programming
Pareto optimization
K-means
Multi-objective Programming
Ideal Point Method
Density-based approach
- Language
In traditional K-means, the target function only considered the intra-cluster similarities and did not take into account the differences among categories. In order to take into account both the firmness in the same cluster and the dispersion between different clusters, a new objective function based on ideal point is given, and the K-means algorithm based on quasi ideal point method is provided. A density-based approach is used to initialize clustering centers. Experimental results show that the improved algorithm obtains more accurate clustering solutions.