Gradient-Sensitive Optimization for Convolutional Neural Networks

Figure 8

Styblinski–Tang function.

(a)

(b)