Module: TLearn

Defined in:
lib/t_learn.rb,
lib/t_learn/em.rb,
lib/t_learn/k_means.rb,
lib/t_learn/version.rb,
lib/t_learn/hop_field_net.rb,
lib/t_learn/feedforward_neural_network.rb

Defined Under Namespace

Classes: EM_Gaussian, FNN, HopFieldNet, K_Means

Constant Summary collapse

VERSION = 
"0.1.1.8"

Class Method Summary collapse

Class Method Details

.add_noise_data(data, noise_rate) ⇒ Object

add noise to sample datas

Parameters:

  • data —

    Array data which we want to add noise

  • noise_rate —

    float rate of noise



103
104
105
106
107
108
109
110
111
112
113
114
115
116
 
# File 'lib/t_learn/hop_field_net.rb', line 103

def TLearn.add_noise_data(data,noise_rate)
  data_with_noise = Marshal.load(Marshal.dump(data))
  data.size.times do |n|
    if rand <= noise_rate
      if data_with_noise[n] == -1.0
        data_with_noise[n] = 1.0
      else
        data_with_noise[n] = -1.0
      end
    end
  end

  return data_with_noise
end

.evaluate(teacher_data, data) ⇒ Object

evaluate predict data with teatcher data



121
122
123
124
125
126
127
128
129
130
 
# File 'lib/t_learn/hop_field_net.rb', line 121

def TLearn.evaluate(teacher_data,data)
  dominator = 0.0
  molecule = 0.0 
  teacher_data.zip(data).each do |td,d|
    dominator += 1 
    molecule += 1 if td == d
  end

  return (molecule/dominator)*100
end