Class: Rumale::Ensemble::AdaBoostClassifier

Inherits:
Base::Estimator show all
Includes:
Base::Classifier
Defined in:
rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb

Overview

AdaBoostClassifier is a class that implements AdaBoost (SAMME.R) for classification. This class uses decision tree for a weak learner.

Reference

  • Zhu, J., Rosset, S., Zou, H., and Hashie, T., “Multi-class AdaBoost,” Technical Report No. 430, Department of Statistics, University of Michigan, 2005.

Examples:

require 'rumale/ensemble/ada_boost_classifier'

estimator =
  Rumale::Ensemble::AdaBoostClassifier.new(
    n_estimators: 10, criterion: 'gini', max_depth: 3, max_leaf_nodes: 10, min_samples_leaf: 5, random_seed: 1)
estimator.fit(training_samples, traininig_labels)
results = estimator.predict(testing_samples)

Instance Attribute Summary collapse

Attributes inherited from Base::Estimator

#params

Instance Method Summary collapse

Methods included from Base::Classifier

#score

Constructor Details

#initialize(n_estimators: 50, criterion: 'gini', max_depth: nil, max_leaf_nodes: nil, min_samples_leaf: 1, max_features: nil, random_seed: nil) ⇒ AdaBoostClassifier

Create a new classifier with AdaBoost.

Parameters:

  • n_estimators (Integer) (defaults to: 50) —

    The numeber of decision trees for contructing AdaBoost classifier.

  • criterion (String) (defaults to: 'gini') —

    The function to evalue spliting point. Supported criteria are ‘gini’ and ‘entropy’.

  • max_depth (Integer) (defaults to: nil) —

    The maximum depth of the tree. If nil is given, decision tree grows without concern for depth.

  • max_leaf_nodes (Integer) (defaults to: nil) —

    The maximum number of leaves on decision tree. If nil is given, number of leaves is not limited.

  • min_samples_leaf (Integer) (defaults to: 1) —

    The minimum number of samples at a leaf node.

  • max_features (Integer) (defaults to: nil) —

    The number of features to consider when searching optimal split point. If nil is given, split process considers all features.

  • random_seed (Integer) (defaults to: nil) —

    The seed value using to initialize the random generator. It is used to randomly determine the order of features when deciding spliting point.



58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 58

def initialize(n_estimators: 50,
               criterion: 'gini', max_depth: nil, max_leaf_nodes: nil, min_samples_leaf: 1,
               max_features: nil, random_seed: nil)
  super()
  @params = {
    n_estimators: n_estimators,
    criterion: criterion,
    max_depth: max_depth,
    max_leaf_nodes: max_leaf_nodes,
    min_samples_leaf: min_samples_leaf,
    max_features: max_features,
    random_seed: random_seed || srand
  }
  @rng = Random.new(@params[:random_seed])
end

Instance Attribute Details

#classes ⇒ Numo::Int32 (readonly)

Return the class labels.

Returns:

  • (Numo::Int32) —

    (size: n_classes)



35
36
37
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 35

def classes
  @classes
end

#estimators ⇒ Array<DecisionTreeClassifier> (readonly)

Return the set of estimators.

Returns:

  • (Array<DecisionTreeClassifier>) 


31
32
33
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 31

def estimators
  @estimators
end

#feature_importances ⇒ Numo::DFloat (readonly)

Return the importance for each feature.

Returns:

  • (Numo::DFloat) —

    (size: n_features)



39
40
41
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 39

def feature_importances
  @feature_importances
end

#rng ⇒ Random (readonly)

Return the random generator for random selection of feature index.

Returns:

  • (Random) 


43
44
45
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 43

def rng
  @rng
end

Instance Method Details

#decision_function(x) ⇒ Numo::DFloat

Calculate confidence scores for samples.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to compute the scores.

Returns:

  • (Numo::DFloat) —

    (shape: [n_samples, n_classes]) Confidence score per sample.



136
137
138
139
140
141
142
143
144
145
146
147
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 136

def decision_function(x)
  x = ::Rumale::Validation.check_convert_sample_array(x)

  n_samples, = x.shape
  n_classes = @classes.size
  sum_probs = Numo::DFloat.zeros(n_samples, n_classes)
  @estimators.each do |tree|
    log_proba = Numo::NMath.log(tree.predict_proba(x).clip(1.0e-15, nil))
    sum_probs += (n_classes - 1) * (log_proba - 1.fdiv(n_classes) * Numo::DFloat[log_proba.sum(axis: 1)].transpose)
  end
  sum_probs /= @estimators.size
end

#fit(x, y) ⇒ AdaBoostClassifier

Fit the model with given training data.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The training data to be used for fitting the model.

  • y (Numo::Int32) —

    (shape: [n_samples]) The labels to be used for fitting the model.

Returns:



79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 79

def fit(x, y) # rubocop:disable Metrics/AbcSize
  x = ::Rumale::Validation.check_convert_sample_array(x)
  y = ::Rumale::Validation.check_convert_label_array(y)
  ::Rumale::Validation.check_sample_size(x, y)

  ## Initialize some variables.
  n_samples, n_features = x.shape
  @estimators = []
  @feature_importances = Numo::DFloat.zeros(n_features)
  @params[:max_features] = n_features unless @params[:max_features].is_a?(Integer)
  @params[:max_features] = [[1, @params[:max_features]].max, n_features].min # rubocop:disable Style/ComparableClamp
  @classes = Numo::Int32.asarray(y.to_a.uniq.sort)
  n_classes = @classes.shape[0]
  sub_rng = @rng.dup
  ## Boosting.
  classes_arr = @classes.to_a
  y_codes = Numo::DFloat.zeros(n_samples, n_classes) - 1.fdiv(n_classes - 1)
  n_samples.times { |n| y_codes[n, classes_arr.index(y[n])] = 1.0 }
  observation_weights = Numo::DFloat.zeros(n_samples) + 1.fdiv(n_samples)
  @params[:n_estimators].times do |_t|
    # Fit classfier.
    ids = ::Rumale::Utils.choice_ids(n_samples, observation_weights, sub_rng)
    break if y[ids].to_a.uniq.size != n_classes

    tree = ::Rumale::Tree::DecisionTreeClassifier.new(
      criterion: @params[:criterion], max_depth: @params[:max_depth],
      max_leaf_nodes: @params[:max_leaf_nodes], min_samples_leaf: @params[:min_samples_leaf],
      max_features: @params[:max_features], random_seed: sub_rng.rand(::Rumale::Ensemble::Value::SEED_BASE)
    )
    tree.fit(x[ids, true], y[ids])
    # Calculate estimator error.
    proba = tree.predict_proba(x).clip(1.0e-15, nil)
    pred = Numo::Int32.asarray(Array.new(n_samples) { |n| @classes[proba[n, true].max_index] })
    inds = pred.ne(y)
    error = (observation_weights * inds).sum / observation_weights.sum
    # Store model.
    @estimators.push(tree)
    @feature_importances += tree.feature_importances
    break if error.zero?

    # Update observation weights.
    log_proba = Numo::NMath.log(proba)
    observation_weights *= Numo::NMath.exp(-1.0 * (n_classes - 1).fdiv(n_classes) * (y_codes * log_proba).sum(axis: 1))
    observation_weights = observation_weights.clip(1.0e-15, nil)
    sum_observation_weights = observation_weights.sum
    break if sum_observation_weights.zero?

    observation_weights /= sum_observation_weights
  end
  @feature_importances /= @feature_importances.sum
  self
end

#predict(x) ⇒ Numo::Int32

Predict class labels for samples.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to predict the labels.

Returns:

  • (Numo::Int32) —

    (shape: [n_samples]) Predicted class label per sample.



153
154
155
156
157
158
159
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 153

def predict(x)
  x = ::Rumale::Validation.check_convert_sample_array(x)

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

#predict_proba(x) ⇒ Numo::DFloat

Predict probability for samples.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to predict the probailities.

Returns:

  • (Numo::DFloat) —

    (shape: [n_samples, n_classes]) Predicted probability of each class per sample.



165
166
167
168
169
170
171
172
173
 
# File 'rumale-ensemble/lib/rumale/ensemble/ada_boost_classifier.rb', line 165

def predict_proba(x)
  x = ::Rumale::Validation.check_convert_sample_array(x)

  n_classes = @classes.size
  probs = Numo::NMath.exp(1.fdiv(n_classes - 1) * decision_function(x))
  sum_probs = probs.sum(axis: 1)
  probs /= Numo::DFloat[sum_probs].transpose
  probs
end