Deep Convolutional Extreme Learning Machine and Its Application in Handwritten Digit Classification

Table 3

Performance on MNIST (with only 15k training samples).

Algorithms	Training time (s)	Training accuracy		Testing accuracy
		Mean	Std	Mean	Std
ELM	169.3	0.9976	0.0012	0.9522	0.0009
LRF-ELM	264.9	0.9890	0.0039	0.9779	0.0021
DC-ELM	120.4	0.9887	0.0020	0.9841	0.0012
CNN	2960.9	0.9873	0.0014	0.9781	0.0011
DBN	4791.3	0.9905	0.0009	0.9759	0.0015