Keras Installation and introduction

Reprint: http://blog.csdn.net/mmc2015/article/details/50976776

Install first and say:

sudo pipinstall Keras

or manually installed:

Download: Git clone git://github.com/fchollet/keras.git

Upload it to the appropriate machine.

Install: CD to the Keras folder and run the Install command:

sudo python setup.py install

Keras in Theano, before learning Keras, first understood this several content:

http://blog.csdn.net/mmc2015/article/details/42222075 (LR)

Http://www.deeplearning.net/tutorial/gettingstarted.html and http://www.deeplearning.net/tutorial/ logreg.html (classifying mnist digits using Logistic regression)

The PLA: http://www.deeplearning.net/tutorial/contents.html

Take the code given in the first link as an example (relatively simple):

[Python]  View plain  copy import numpy   import theano   import theano.tensor  as T   rng = numpy.random      n = 400                                     # training  sample size   feats = 784                                 # number of input variables      # generate  a dataset: D =  (Input_values, target_class)    d =  (rng.randn (N, feats),  rng.randint (size=n, low=0, high=2))    training_steps =  10000 &nBsp    # declare theano symbolic variables   X = t.matrix ("x")    y = t.vector ("y")       # initialize the weight  vector w randomly   #   # this and the following bias  variable b   # are shared so they keep their values    # between training iterations  (Updates)    w = theano.shared (Rng.randn (feats),  name= "W")       # initialize the bias term    B = theano.shared (0., name= "B")       print ("Initial model:")     Print (W.get_value ())    print (B.get_value ())       # construct theano  expression graph   p_1 = 1 /  (1 + t.exp-t.dot (x, w) & NBsP;- B))    # Probability that target = 1   prediction  = p_1 > 0.5                     # The prediction thresholded   xent  = -y * t.log (p_1)  -  (1-y)  * t.log (1-p_1)  # cross-entropy  loss function   Cost = xent.mean ()  + 0.01 *  (w **  2. SUM () # the cost to minimize   Gw, gb = t.grad (Cost, [w,  B])              # compute the  gradient of the cost                                    &nBsp;          # w.r.t weight vector w  and                                               # bias term b                                                #  (we shall return to this in a