Skip to main content
. 2019 Apr 11;9(2):33–44. doi: 10.1093/af/vfz005

Figure 7.

Figure 7.

A schematic representation of how (A) classical programming and (B) “learning” programming paradigms use inputs, outputs, and codes (i.e., rules) for predictions. The arrow from code to “learning” programming indicates back propagation frequently used in deep learning. Adapted from Chollet (2018).