Table 4 The average accuracy of and time needed by Q and C neural networks with different numbers of neurons for three real datasets. The notation \((\cdot )\) stands for the architecture of a neural network as described in Eq. 15
| Â | Yeast | Pendigits | Isolet | |||
|---|---|---|---|---|---|---|
|  | Accuracy (%) | Time (s) | Accuracy (%) | Time (s) | Accuracy (%) | Time (s) |
C(2-3) | \(57.17 \pm 1.79\) | \(0.5 \pm 0.3\) | \(90.01 \pm 1.54\) | \(3.3 \pm 2.7\) | \(89.94 \pm 3.30\) | \(11.2 \pm 0.2\) |
C(2-10-3) | \(58.21 \pm 1.93\) | \(0.7 \pm 0.2\) | \(93.23 \pm 2.39\) | \(4.0 \pm 1.9\) | \(92.00 \pm 0.60\) | \(11.8 \pm 0.7\) |
C(2-100-3) | \(59.69 \pm 2.96\) | \(0.9 \pm 0.2\) | \(96.66 \pm 0.29\) | \(14.8 \pm 27.4\) | \(94.47 \pm 0.30\) | \(49.8 \pm 2.6\) |
Q(2-3) | \(60.99 \pm 1.27\) | \(0.5 \pm 0.1\) | \(97.04 \pm 0.30\) | \(6.1 \pm 0.2\) | \(95.01 \pm 0.17\) | \(21.2 \pm 0.2\) |