scikit-learn：如何在由一个列表组成的嵌套列表上使用 RandomizedSearchCV？

Question

I have built a Sentence Boundary Detection Classifier. For the sequence labeling I used a conditional random field. For the hyperparameter optimization I would like to use RandomizedSearchCV. My training data consists of 6 annotated texts. I merge all 6 texts to a tokenlist. For the implementation I followed an example from the documentation . Here my simplified code:

from sklearn_crfsuite import CRF
from sklearn_crfsuite import metrics
from sklearn.metrics import make_scorer
from sklearn.model_selection import RandomizedSearchCV
from sklearn.model_selection import train_test_split
import scipy.stats

#my tokenlist has the length n
X_train = [feature_dict_token_1, ... , feature_dict_token_n]
# 3 types of tags, B-SEN for begin of sentence; E-SEN for end of sentence; O-Others
y_train = [tag_token_1, ..., tag_token_n]

# define fixed parameters and parameters to search
crf = sklearn_crfsuite.CRF(
    algorithm='lbfgs',
    max_iterations=100,
    all_possible_transitions=True
)
params_space = {
    'c1': scipy.stats.expon(scale=0.5),
    'c2': scipy.stats.expon(scale=0.05),
}

labels = ['B-SEN', 'E-SEN', 'O']

# use F1-score for evaluation
f1_scorer = make_scorer(metrics.flat_f1_score,
                        average='weighted', labels=labels)

# search
rs = RandomizedSearchCV(crf, params_space,
                        cv=3,
                        verbose=1,
                        n_jobs=-1,
                        n_iter=50,
                        scoring=f1_scorer)
rs.fit([X_train], [y_train])

I used rs.fit([X_train], [y_train]) instead of rs.fit(X_train, y_train) since the documentation of crf.train says, that it needs a list of lists:

fit(X, y, X_dev=None, y_dev=None)

Parameters: 
-X (list of lists of dicts) – Feature dicts for several documents (in a python-crfsuite format).
-y (list of lists of strings) – Labels for several documents.
-X_dev ((optional) list of lists of dicts) – Feature dicts used for testing.
-y_dev ((optional) list of lists of strings) – Labels corresponding to X_dev.

But using a list of lists I get this Error:

ValueError: Cannot have number of splits n_splits=5 greater than the number of samples: n_samples=1

I understand that it is because I use [X_train] and [y_train] respectively and it is not possible to apply CV to a list consisting of one list, but with X_train and y_train crf.fit does not cope. How can i fix this?

Answer 1

According to the official tutorialhere , your train/test sets (ie, X_train , X_test ) should be a list of lists of dictionaries. For example:

[[{'bias': 1.0,
   'word.lower()': 'melbourne',
   'word[-3:]': 'rne',
   'word[-2:]': 'ne',
   'word.isupper()': False,
   'word.istitle()': True,
   'word.isdigit()': False,
   'postag': 'NP'},
  {'bias': 1.0,
   'word.lower()': '(',
   'word[-3:]': '(',
   'word[-2:]': '(',
   'word.isupper()': False,
   'word.istitle()': False,
   'word.isdigit()': False,
   'postag': 'Fpa'},
   ...],
    [{'bias': 1.0,
   'word.lower()': '-',
   'word[-3:]': '-',
   'word[-2:]': '-',
   'word.isupper()': False,
   'word.istitle()': False,
   'word.isdigit()': False,
   'postag': 'Fg',
   'postag[:2]': 'Fg'},
    {'bias': 1.0,
   'word.lower()': '25',
   'word[-3:]': '25',
   'word[-2:]': '25',
   'word.isupper()': False,
   'word.istitle()': False,
   'word.isdigit()': True,
   'postag': 'Z'
   }]]

The labels sets (ie, y_tain and y_test) should be a list of lists of strings. For instance:

[['B-LOC', 'I-LOC'], ['B-ORG', 'O']]

Then you fit the model as normally:

rs.fit(X_train, y_train)

Please take the tutorial mentioned above to see how that works.