Neural network tutorial, softmax activation instead of sigmoid

Question

I am working through an FNN tutorial, previously after researching i learnt that i will need to use softmax activation for my own ML problem rather than sigmoid as shown.

After hours of looking through tutorials i cannot find a basic example of softmax code (outside of modules) that i can learn to rework the tutorial code in the way that sigmoid is used. If i can figure this out then i can break it down and understand how the math transforms the data and is fed through the NN, then i can apply this to other basic ML starting points like SVMs etc.

I need pointers on untangling the sigmoid math/code and reworking it to use softmax activation and one hot encoding on the y outputs, thank you for any help.

import numpy as np

# sigmoid function
def nonlin(x,deriv=False):
    if(deriv==True):
        return x*(1-x)
    return 1/(1+np.exp(-x))

# input dataset
X = np.array([  [0,0,1],
                [0,1,1],
                [1,0,1],
                [1,1,1] ])

# output dataset            
y = np.array([[0,0,1,1]]).T

# seed random numbers to make calculation
# deterministic (just a good practice)
np.random.seed(1)

# initialize weights randomly with mean 0
syn0 = 2*np.random.random((3,1)) - 1

for iter in xrange(10000):

    # forward propagation
    l0 = X
    l1 = nonlin(np.dot(l0,syn0))

    # how much did we miss?
    l1_error = y - l1

    # multiply how much we missed by the 
    # slope of the sigmoid at the values in l1
    l1_delta = l1_error * nonlin(l1,True)

    # update weights
    syn0 += np.dot(l0.T,l1_delta)

print "Output After Training:"
print l1

Neural network tutorial, softmax activation instead of sigmoid

0 Answers0