A mutual information based online evolving clustering approach and its applications
2017
In this article, a new recursive evolving clustering method is proposed based on the well-known Gustafson–Kessel algorithm. The novelty of the proposed method involves the adaptation and integration of the mutual information based formulation to accommodate the Mahalanobis distance, which functions as the similarity measure and the unification of the clustering generation and pruning mechanisms. Example applications of the method are also discussed in the areas of data compression and knowledge extraction.
Keywords:
- Correlation clustering
- Machine learning
- Artificial intelligence
- Pattern recognition
- FLAME clustering
- Cluster analysis
- Computer science
- Data stream clustering
- Conceptual clustering
- Canopy clustering algorithm
- Brown clustering
- CURE data clustering algorithm
- Clustering high-dimensional data
- Fuzzy clustering
- Data mining
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
37
References
1
Citations
NaN
KQI