Skip to main content
. 2021 Aug 20;7:e679. doi: 10.7717/peerj-cs.679

Table 2. Datasets.

Dataset k n d
Blobs 3 1,000 2
Circles 2 1,000 2
Moons 2 1,000 2
Iris 3 150 4
Wine 3 178 13
Spam 2 4,601 54
CNAE-9 9 1,080 856
Digits 10 1,797 8 × 8
MNIST 10 70,000 28 × 28

Note:

k, n, and d indicate the number of classes, data points and attributions, respectively.