Class: Ai4r::Classifiers::MultilayerPerceptron

Inherits:
Classifier
  • Object
show all
Defined in:
lib/ai4r/classifiers/multilayer_perceptron.rb

Overview

Introduction

The idea behind the MultilayerPerceptron classifier is to train a Multilayer Perceptron neural network with the provided examples, and predict the class for new data items.

Parameters

Use class method get_parameters_info to obtain details on the algorithm parameters. Use set_parameters to set values for this parameters. See Parameterizable module documentation.

  • :network_class => Neural network implementation class. By default: Ai4r::NeuralNetwork::Backpropagation.

  • :network_parameters => Parameters to be forwarded to the back end neural ntework.

  • :hidden_layers => Hidden layer structure. E.g. [8, 6] will generate 2 hidden layers with 8 and 6 neurons each. By default []

  • :training_iterations => How many times the training should be repeated. By default: 500.

:active_node_value => Default: 1 :inactive_node_value => Default: 1

Constant Summary collapse

TRAINING_ITERATIONS = 
500

Instance Attribute Summary collapse

Instance Method Summary collapse

Methods included from Data::Parameterizable

#get_parameters, included, #set_parameters

Constructor Details

#initialize ⇒ Object 



57
58
59
60
61
62
63
64
65
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 57

def initialize
  super()
  @network_class = Ai4r::NeuralNetwork::Backpropagation
  @hidden_layers = []
  @training_iterations = TRAINING_ITERATIONS
  @network_parameters = {}
  @active_node_value = 1
  @inactive_node_value = 0
end

Instance Attribute Details

#class_value ⇒ Object (readonly)

Returns the value of attribute class_value.



41
42
43
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 41

def class_value
  @class_value
end

#data_set ⇒ Object (readonly)

Returns the value of attribute data_set.



41
42
43
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 41

def data_set
  @data_set
end

#domains ⇒ Object (readonly)

Returns the value of attribute domains.



41
42
43
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 41

def domains
  @domains
end

#network ⇒ Object (readonly)

Returns the value of attribute network.



41
42
43
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 41

def network
  @network
end

Instance Method Details

#build(data_set) ⇒ Object

Build a new MultilayerPerceptron classifier. You must provide a DataSet instance as parameter. The last attribute of each item is considered as the item class.

Parameters:

  • data_set (Object)

Returns:

  • (Object) 


72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 72

def build(data_set)
  data_set.check_not_empty
  @data_set = data_set
  @domains = @data_set.build_domains.collect(&:to_a)
  @outputs = @domains.last.length
  @inputs = 0
  @domains[0...-1].each { |domain| @inputs += domain.length }
  @structure = [@inputs] + @hidden_layers + [@outputs]
  @network = @network_class.new @structure
  inputs = []
  outputs = []
  data_set.data_items.each do |data_item|
    inputs << data_to_input(data_item[0...-1])
    outputs << data_to_output(data_item.last)
  end
  @network.train_epochs(inputs, outputs,
                        epochs: @training_iterations, batch_size: 1)
  self
end

#eval(data) ⇒ Object

You can evaluate new data, predicting its class. e.g.

classifier.eval(['New York',  '<30', 'F'])  # => 'Y'

Parameters:

  • data (Object)

Returns:

  • (Object) 


98
99
100
101
102
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 98

def eval(data)
  input_values = data_to_input(data)
  output_values = @network.eval(input_values)
  @domains.last[get_max_index(output_values)]
end

#get_rules ⇒ Object

Multilayer Perceptron Classifiers cannot generate human-readable rules.

Returns:

  • (Object) 


107
108
109
 
# File 'lib/ai4r/classifiers/multilayer_perceptron.rb', line 107

def get_rules
  "raise 'Neural networks classifiers do not generate human-readable rules.'"
end