data_mining:neural_network:debugging

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
data_mining:neural_network:debugging [2017/08/20 00:08] – phreazerdata_mining:neural_network:debugging [2017/08/20 00:12] (current) – phreazer
Line 14: Line 14:
  
 Take $dW^{[1]}, db^{[1]}, \dots, dW^{[L]},db^{[L]}$ and put it in a big vector $d\theta$. Take $dW^{[1]}, db^{[1]}, \dots, dW^{[L]},db^{[L]}$ and put it in a big vector $d\theta$.
 +
 +J is now $J(\Theta) = J(\Theta_1, ...)$
 +
 +For each i:
 +
 +$d\Theta_{approx}[i] = \frac{J(\dots, \Theta_i+\epsilon,\dots) - J(\dots, \Theta_i-\epsilon,\dots)}{2\epsilon}$
 +
 +$\epsilon = 10^{-7}$
  • data_mining/neural_network/debugging.1503180513.txt.gz
  • Last modified: 2017/08/20 00:08
  • by phreazer