Novel Group Detection and Analysis Method Based on Automatic and Fast Density Clustering.
Ke JinPeng BaoWeiwei XingPublished in: DMS (2016)
Keyphrases
- clustering method
- false positive rate
- detection method
- spectral analysis
- fully automatic
- cost function
- similarity function
- high precision
- anomaly detection
- correspondence analysis
- hierarchical clustering
- spectral clustering
- detection rate
- false positives
- detection algorithm
- unsupervised learning
- experimental evaluation
- dynamic programming
- computational complexity
- preprocessing
- pairwise
- k means
- genetic algorithm
- input data
- features extraction
- similarity measure
- objective function
- high dimensional
- cluster analysis
- significant improvement
- semi automatic
- segmentation method
- object detection
- high accuracy
- data sets
- mutual information