What is SoftmaxLayer in PyBrain

Recipe Objective - What is SoftmaxLayer in PyBrain?

Pybrain provides mainly two layers: TanhLayer and SoftmaxLayer. SoftmaxLayer implements a softmax distribution over the input.

For more related projects -

https://www.projectpro.io/projects/data-science-projects/deep-learning-projects
https://www.projectpro.io/projects/data-science-projects/keras-deep-learning-projects

Let's try to build a network using SoftmaxLayer -

# Importing libraries
from pybrain.tools.shortcuts import buildNetwork
from pybrain.structure import SoftmaxLayer
from pybrain.datasets import SupervisedDataSet
from pybrain.supervised.trainers import BackpropTrainer

# Create a network with two inputs, two hidden, and one output
network = buildNetwork(2, 2, 1, bias=True, hiddenclass=SoftmaxLayer)

# Create a dataset that matches network input and output sizes:
and_gate = SupervisedDataSet(2, 1)

# Create a dataset to be used for testing.
and_testing_data = SupervisedDataSet(2, 1)

# Add input and target values to dataset values for AND truth table
and_gate.addSample((0, 0), (0,))
and_gate.addSample((0, 1), (0,))
and_gate.addSample((1, 0), (0,))
and_gate.addSample((1, 1), (1,))

# Add input and target values to dataset values for AND truth table
and_testing_data.addSample((0, 0), (0,))
and_testing_data.addSample((0, 1), (0,))
and_testing_data.addSample((1, 0), (0,))
and_testing_data.addSample((1, 1), (1,))

#Training the network with dataset and_gate.
backprop_trainer = BackpropTrainer(network, and_gate)

# 5000 iteration on training data.
for iteration in range(5000):
   backprop_trainer.train()

# Testing data
backprop_trainer.testOnData(dataset=and_testing_data, verbose = True)


Output -
Testing on data:
('out:    ', '[-0.008]')
('correct:', '[0     ]')
error:  0.00003074
('out:    ', '[0.334 ]')
('correct:', '[0     ]')
error:  0.05586374
('out:    ', '[0.334 ]')
('correct:', '[0     ]')
error:  0.05586374
('out:    ', '[0.334 ]')
('correct:', '[1     ]')
error:  0.22160712
('All errors:', [3.07437732734914e-05, 0.055863744147581454, 0.05586374414543854, 0.22160712372514815])
('Average error:', 0.08334133894786042)
('Max error:', 0.22160712372514815, 'Median error:', 0.055863744147581454)
0.08334133894786042

In this way, we can use SoftmaxLayer to build a network in pybrain.

​