Class: SvmToolkit::Svm::SvmTrainer

Inherits:
RecursiveTask
  • Object
show all
Defined in:
lib/svm_toolkit/svm.rb

Overview

Represent a single training task for an SVM RBF model

Instance Method Summary collapse

Constructor Details

#initialize(training_set, parameters, cross_valn_set, evaluator) ⇒ SvmTrainer

Creates an instance of an SvmTrainer.

  • training_set - used to train the model

  • parameters - parameters for building the model

  • cross_valn_set - used to test the model performance

  • evaluator - class name of Evaluator to use for evaluating the model performance



130
131
132
133
134
135
136
137
 
# File 'lib/svm_toolkit/svm.rb', line 130

def initialize training_set, parameters, cross_valn_set, evaluator
  super()

  @training_set = training_set
  @parameters = parameters
  @cross_valn_set = cross_valn_set
  @evaluator = evaluator
end

Instance Method Details

#compute ⇒ Object

Trains and evaluates a model, using the parameters.

Returns the model and evaluation.



143
144
145
146
147
 
# File 'lib/svm_toolkit/svm.rb', line 143

def compute
  model = Svm.svm_train @training_set, @parameters
  result = model.evaluate_dataset @cross_valn_set, :evaluator => @evaluator
  return model, result
end