Class: SVMKit::Multiclass::OneVsRestClassifier

Inherits:
Object
  • Object
show all
Includes:
Base::BaseEstimator, Base::Classifier
Defined in:
lib/svmkit/multiclass/one_vs_rest_classifier.rb

Overview

OneVsRestClassifier is a class that implements One-vs-Rest (OvR) strategy for multi-label classification.

Examples:

base_estimator =
 SVMKit::LinearModel::PegasosSVC.new(penalty: 1.0, max_iter: 100, batch_size: 20, random_seed: 1)
estimator = SVMKit::Multiclass::OneVsRestClassifier.new(estimator: base_estimator)
estimator.fit(training_samples, training_labels)
results = estimator.predict(testing_samples)

Instance Attribute Summary collapse

Attributes included from Base::BaseEstimator

#params

Instance Method Summary collapse

Constructor Details

#initialize(estimator: nil) ⇒ OneVsRestClassifier

Create a new multi-label classifier with the one-vs-rest startegy.



30
31
32
33
34
35
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 30

def initialize(estimator: nil)
  @params = {}
  @params[:estimator] = estimator
  @estimators = nil
  @classes = nil
end

Instance Attribute Details

#classes ⇒ Numo::Int32 (readonly)

Return the class labels.



25
26
27
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 25

def classes
  @classes
end

#estimators ⇒ Array<Classifier> (readonly)

Return the set of estimators.



21
22
23
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 21

def estimators
  @estimators
end

Instance Method Details

#decision_function(x) ⇒ Numo::DFloat

Calculate confidence scores for samples.



56
57
58
59
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 56

def decision_function(x)
  n_classes = @classes.size
  Numo::DFloat.asarray(Array.new(n_classes) { |m| @estimators[m].decision_function(x).to_a }).transpose
end

#fit(x, y) ⇒ OneVsRestClassifier

Fit the model with given training data.



42
43
44
45
46
47
48
49
50
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 42

def fit(x, y)
  y_arr = y.to_a
  @classes = Numo::Int32.asarray(y_arr.uniq.sort)
  @estimators = @classes.to_a.map do |label|
    bin_y = Numo::Int32.asarray(y_arr.map { |l| l == label ? 1 : -1 })
    @params[:estimator].dup.fit(x, bin_y)
  end
  self
end

#marshal_dump ⇒ Hash

Dump marshal data.



84
85
86
87
88
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 84

def marshal_dump
  { params: @params,
    classes: @classes,
    estimators: @estimators.map { |e| Marshal.dump(e) } }
end

#marshal_load(obj) ⇒ nil

Load marshal data.



92
93
94
95
96
97
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 92

def marshal_load(obj)
  @params = obj[:params]
  @classes = obj[:classes]
  @estimators = obj[:estimators].map { |e| Marshal.load(e) }
  nil
end

#predict(x) ⇒ Numo::Int32

Predict class labels for samples.



65
66
67
68
69
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 65

def predict(x)
  n_samples, = x.shape
  decision_values = decision_function(x)
  Numo::Int32.asarray(Array.new(n_samples) { |n| @classes[decision_values[n, true].max_index] })
end

#score(x, y) ⇒ Float

Claculate the mean accuracy of the given testing data.



76
77
78
79
80
 
# File 'lib/svmkit/multiclass/one_vs_rest_classifier.rb', line 76

def score(x, y)
  p = predict(x)
  n_hits = (y.to_a.map.with_index { |l, n| l == p[n] ? 1 : 0 }).inject(:+)
  n_hits / y.size.to_f
end