Skip to main content

Advertisement

Table 1 Results of the application of our proposed algorithm (PA) and of the BFR algorithm to synthetic data

From: A clustering algorithm for multivariate data streams with correlated components

N. of true clusters Algorithm Dimension p of data points N. of data in each chunk N. of estimated clusters N. of retained points (outliers)
5 BFR 5 25 6 0
5 PA 5 25 5 0
5 BFR 5 50 6 0
5 PA 5 50 5 0
5 BFR 10 25 5 0
5 PA 10 25 5 0
5 BFR 10 50 5 0
5 PA 10 50 5 0
5 BFR 20 25 5 0
5 PA 20 25 5 0
5 BFR 20 50 5 0
5 PA 20 50 5 0
20 BFR 10 25 12 0
20 PA 10 25 17 0
20 BFR 10 50 13 0
20 PA 10 50 22 1
20 BFR 20 25 11 0
20 PA 20 25 19 0
20 BFR 20 50 20 0
20 PA 20 50 20 0
  1. We call chunk the number of processed data out of which we apply secondary compression