sample config

config.yml

models:
  1:
    vectorizer:
      title:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1], [1, 2]]
            max_features: [500]
      url:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
    estimator:
      name: 'SVC'
      parameters:
        - kernel: ['rbf']
          gamma: [1.0e-3, 1.0e-4]
          C: [1, 10, 100, 1000]
        - kernel: ['linear']
          C: [1, 10, 100, 1000]
  2:
    vectorizer:
      title:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
      url:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
    estimator:
      name: 'MultinomialNB'
      parameters:
        - alpha: [1.0, 2.0, 5.0, 10.0, 20.0, 25.0]
  3:
    vectorizer:
      title:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
      url:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
    estimator:
      name: 'LogisticRegression'
      parameters:
      - class_weight:
        - 0: 0.8
          1: 0.2
        - 0: 0.85
          1: 0.15
        - 0: 0.9
          1: 0.1
  4:
    vectorizer:
      title:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
      url:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
    estimator:
      name: 'Perceptron'
      parameters:
        - alpha: [1.0e-4, 1.0e-3]
          max_iter: [5]
  5:
    vectorizer:
      title:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
      url:
        name: 'tfidf'
        parameters:
          - ngram_range: [[1, 1]]
            max_features: [500]
    estimator:
      name: 'MLPClassifier'
      parameters:
        - activation: ['relu', 'logistic', 'tanh']
          max_iter: [200]
cross_validation:
    train_test_split: [75, 25] # these need to sum to 100
    n_folds: 5
test:
  scores: ['precision', 'recall', 'accuracy']
data: 'data/data.csv'