DP-Dip: A skinny method for estimating the number and center of clusters.
Shuaijing XuRongfang BieLiangchi LiYuqi YangPublished in: IIKI (2017)
Keyphrases
- dynamic programming
- computational cost
- detection method
- synthetic data
- small number
- experimental evaluation
- classification accuracy
- initial set
- hierarchical clustering
- detection algorithm
- high accuracy
- neural network
- prior knowledge
- preprocessing
- computational complexity
- training data
- significant improvement
- high dimensional
- training set
- high precision
- data clustering
- cluster centers