Sparse Data Analysis Strategy for Neural Spike Classification

Table 4

Comparison of 3 methods of seriation, PB-Clus, HC, and Chen approach according to Moore and Neumann criteria on the data benchmarks.

Method	PB-Clus seriation		HC seriation		Chen seriation
Criterion	Moore	Neumann	Moore	Neumann	Moore	Neumann
Dataset
Iris	1 371.2	471.1	31 728.8	10 893.1	19 357.8	7 304.0
Townships	244.5	91.8	1 109.9	441.5	849.0	342.0
Ruspini	1 290.1	442.2	8 724.9	3 036.4	6 503.7	2 277.1
Faithful	2 634.1	889.4	34 045.5	11 503.5	23 390.0	9 894.2
Geysers	2 514.9	850.4	68 205.3	2 302.1	12 866.8	4 501.4