Convolutional Neural Networks for Recognition of Lymphoblast Cell Images

Table 5

Accuracy of ConVNet, SVM-GA, MLP, and random forest to identify lymphocytes, pre-T, and pre-B cells over ten test sets.

Test set	ConVNet	SVM-GA	MLP	Random forest
1	82.64	81.82	74.38	78.51
2	80.17	80.99	75.21	72.73
3	85.12	80.17	75.21	79.34
4	80.17	80.17	76.86	80.99
5	78.51	79.34	76.03	78.51
6	78.51	81.82	73.55	76.86
7	83.47	80.99	78.51	79.34
8	85.95	81.82	81.82	80.99
9	83.47	86.78	76.03	79.34
10	79.34	82.64	73.55	77.69
Average	81.74 ± 2.74	81.65 ± 2.05	76.12 ± 2.51	78.43 ± 2.38