Ideas for Neural Network

  • Use genetic algorithm to evolve the parameters in Neural Network, to make a self-develop algorithm.
    • automatically find and make training set from search engine
    • the paremeters include:
      • learning rate
      • iteration times
      • regularization term \(\lambda\)
      • activation function
      • number of hidden layers(units)
Read more   2016/3/4 20:44 下午 posted in  Pattern Recognition

Notes for Neural Network

Notes:

  • If we set the initial \(\Theta\) be the same, the units in next layer with the same \(x_i\) will get the same result, then all units in the same layer will get the same output. At last, the cost function will also get same cost, so we will update the \(\Theta\) with same step.

  • seems \(\delta^{(l)}_{i}\) means the cost of \(i_{th}\) unit in the \(l_{th}\) layer
Read more   2016/3/3 0:57 上午 posted in  Pattern Recognition

Copyright © 2016 Powered by MWeb,  Theme used GitHub CSS. Modified by Yinzo.