Class: Rumale::SVM::OneClassSVM

Inherits:
Base::Estimator
  • Object
show all
Defined in:
lib/rumale/svm/one_class_svm.rb

Overview

OneClassSVM is a class that provides One-class Support Vector Machine in LIBSVM with Rumale interface.

Examples:

estimator = Rumale::SVM::OneClassSVM.new(nu: 0.5, kernel: 'rbf', gamma: 10.0, random_seed: 1)
estimator.fit(training_samples, traininig_labels)
results = estimator.predict(testing_samples)

Instance Method Summary collapse

Constructor Details

#initialize(nu: 1.0, kernel: 'rbf', degree: 3, gamma: 1.0, coef0: 0.0, shrinking: true, probability: true, cache_size: 200.0, tol: 1e-3, verbose: false, random_seed: nil) ⇒ OneClassSVM

Create a new estimator with One-class Support Vector Machine.

Parameters:

  • nu (Float) (defaults to: 1.0) —

    The regularization parameter. The interval of nu is (0, 1].

  • kernel (String) (defaults to: 'rbf') —

    The type of kernel function (‘rbf’, ‘linear’, ‘poly’, ‘sigmoid’, and ‘precomputed’).

  • degree (Integer) (defaults to: 3) —

    The degree parameter in polynomial kernel function.

  • gamma (Float) (defaults to: 1.0) —

    The gamma parameter in rbf/poly/sigmoid kernel function.

  • coef0 (Float) (defaults to: 0.0) —

    The coefficient in poly/sigmoid kernel function.

  • shrinking (Boolean) (defaults to: true) —

    The flag indicating whether to use the shrinking heuristics.

  • probability (Boolean) (defaults to: true) —

    The flag indicating whether to train the parameter for probability estimation.

  • cache_size (Float) (defaults to: 200.0) —

    The cache memory size in MB.

  • tol (Float) (defaults to: 1e-3) —

    The tolerance of termination criterion.

  • verbose (Boolean) (defaults to: false) —

    The flag indicating whether to output learning process message

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

    The seed value using to initialize the random generator.



29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
 
# File 'lib/rumale/svm/one_class_svm.rb', line 29

def initialize(nu: 1.0, kernel: 'rbf', degree: 3, gamma: 1.0, coef0: 0.0,
               shrinking: true, probability: true, cache_size: 200.0, tol: 1e-3, verbose: false, random_seed: nil)
  super()
  @params = {}
  @params[:nu] = nu.to_f
  @params[:kernel] = kernel
  @params[:degree] = degree.to_i
  @params[:gamma] = gamma.to_f
  @params[:coef0] = coef0.to_f
  @params[:shrinking] = shrinking
  @params[:probability] = probability
  @params[:cache_size] = cache_size.to_f
  @params[:tol] = tol.to_f
  @params[:verbose] = verbose
  @params[:random_seed] = random_seed.nil? ? nil : random_seed.to_i
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. If the kernel is ‘precomputed’, the shape of x must be [n_samples, n_training_samples].

Returns:

  • (Numo::DFloat) —

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



64
65
66
67
68
 
# File 'lib/rumale/svm/one_class_svm.rb', line 64

def decision_function(x)
  raise "#{self.class.name}##{__method__} expects to be called after training the model with the fit method." unless trained?
  x = Rumale::Validation.check_convert_sample_array(x)
  Numo::Libsvm.decision_function(x, libsvm_params, @model)
end

#duel_coef ⇒ Numo::DFloat

Return the coefficients of the support vector in decision function.

Returns:

  • (Numo::DFloat) —

    (shape: [1, n_support_vectors])



129
130
131
 
# File 'lib/rumale/svm/one_class_svm.rb', line 129

def duel_coef
  @model[:sv_coef]
end

#fit(x) ⇒ OneClassSVM

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. If the kernel is ‘precomputed’, x must be a square distance matrix (shape: [n_samples, n_samples]).

Returns:



52
53
54
55
56
57
 
# File 'lib/rumale/svm/one_class_svm.rb', line 52

def fit(x, _y = nil)
  x = Rumale::Validation.check_convert_sample_array(x)
  dummy = Numo::DFloat.ones(x.shape[0])
  @model = Numo::Libsvm.train(x, dummy, libsvm_params)
  self
end

#intercept ⇒ Numo::DFloat

Return the intercepts in decision function.

Returns:

  • (Numo::DFloat) —

    (shape: [1])



135
136
137
 
# File 'lib/rumale/svm/one_class_svm.rb', line 135

def intercept
  @model[:rho]
end

#marshal_dump ⇒ Hash

Dump marshal data.

Returns:

  • (Hash) —

    The marshal data about SVC.



96
97
98
99
 
# File 'lib/rumale/svm/one_class_svm.rb', line 96

def marshal_dump
  { params: @params,
    model: @model }
end

#marshal_load(obj) ⇒ nil

Load marshal data.

Returns:

  • (nil) 


103
104
105
106
107
 
# File 'lib/rumale/svm/one_class_svm.rb', line 103

def marshal_load(obj)
  @params = obj[:params]
  @model = obj[:model]
  nil
end

#n_support ⇒ Integer

Return the number of support vectors.

Returns:

  • (Integer) 


123
124
125
 
# File 'lib/rumale/svm/one_class_svm.rb', line 123

def n_support
  @model[:sv_indices].size
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. If the kernel is ‘precomputed’, the shape of x must be [n_samples, n_training_samples].

Returns:

  • (Numo::Int32) —

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



75
76
77
78
79
 
# File 'lib/rumale/svm/one_class_svm.rb', line 75

def predict(x)
  raise "#{self.class.name}##{__method__} expects to be called after training the model with the fit method." unless trained?
  x = Rumale::Validation.check_convert_sample_array(x)
  Numo::Int32.cast(Numo::Libsvm.predict(x, libsvm_params, @model))
end

#predict_proba(x) ⇒ Numo::DFloat

Predict class probability for samples. This method works correctly only if the probability parameter is true.

Parameters:

  • x (Numo::DFloat) —

    (shape: [n_samples, n_features]) The samples to predict the probailities. If the kernel is ‘precomputed’, the shape of x must be [n_samples, n_training_samples].

Returns:

  • (Numo::DFloat) —

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



87
88
89
90
91
92
 
# File 'lib/rumale/svm/one_class_svm.rb', line 87

def predict_proba(x)
  raise "#{self.class.name}##{__method__} expects to be called after training the model with the fit method." unless trained?
  raise "#{self.class.name}##{__method__} expects to be called after training the probablity parameters." unless trained_probs?
  x = Rumale::Validation.check_convert_sample_array(x)
  Numo::Libsvm.predict_proba(x, libsvm_params, @model)
end

#support ⇒ Numo::Int32

Return the indices of support vectors.

Returns:

  • (Numo::Int32) —

    (shape: [n_support_vectors])



111
112
113
 
# File 'lib/rumale/svm/one_class_svm.rb', line 111

def support
  @model[:sv_indices]
end

#support_vectors ⇒ Numo::DFloat

Return the support_vectors.

Returns:

  • (Numo::DFloat) —

    (shape: [n_support_vectors, n_features])



117
118
119
 
# File 'lib/rumale/svm/one_class_svm.rb', line 117

def support_vectors
  @model[:SV]
end