[keghn]

Unsupervised Representation Learning:

[yotamarker]

can someone explain 46:30

[keghn]

 The low point is where a trainable algorithm give the correct answer of y2 at y1. But correct answer change when y1 changes.

[keghn]

A Neural Network in 13 lines of Python (Part 2 - Gradient Descent): 

http://iamtrask.github.io/2015/07/27/python-network-part2/