Figure 4. Numerical and experimental results for spoken digit recognition.

The y-axis on the left-hand side denotes the margin, whereas the y-axis on the right-hand side denotes the word error rate. The abscissa represents the parameter η. γ has been kept fixed at 0.5 and the exponent is set to p=7. The delay time is set at τ=80, with N=400 neurons of θ=0.2 separation. The red line represents results for the numerically obtained margin and the black line represents the numerically obtained word error rate. The red and black crosses denote the corresponding experimental results.