from keras.models import Model
from keras.wrappers.scikit_learn import KerasClassifier
from run_multilabel_classifier import run
def create_model():
inp = Input(shape=())
embed_size = 128
x = Embedding(2000, embed_size)(inp)
x = LSTM(60, return_sequences=True, name='lstm_layer')(inp)
x = GlobalMaxPool1D()(x)
x = Dropout(0.1)(x)
x = Dense(50, activation="relu")(x)
x = Dropout(0.1)(x)
x = Dense(1, activation="sigmoid")(x)
model = Model(inputs=inp, outputs=x)
model.compile(loss='binary_crossentropy',
optimizer='adam',
metrics=['accuracy'])
return model
param_grid = {'estimator__classifier__C': [100., 0.1]}
estimator = KerasClassifier(build_fn=create_model,
epochs=1,
batch_size=16,
verbose=1)
run(param_grid, estimator)
# =========================== #
# TRAIN
# BINARY CLASSIFIER
# =========================== #
binary_clf = run_binary_classifier.run(binary_param_grid, LogisticRegression(), comments_file=train_binary)
with open('./saved_models/log_reg_joint_binary.pkl', 'wb') as saved_model:
pickle.dump(binary_clf, file=saved_model)
# =========================== #
# TRAIN
# MULTILABEL CLASSIFIER
# =========================== #
multilabel_clf = run_multilabel_classifier.run(multilabel_param_grid, LogisticRegression(), comments_file=train_multilabel)
with open('./saved_models/log_reg_joint_multilabel.pkl', 'wb') as saved_model:
pickle.dump(binary_clf, file=saved_model)
# =========================== #
# PREDICT
# BINARY CLASSIFIER
# =========================== #
print('Binary prediction')
X_binary_test, y_binary_test = run_binary_classifier.load_comments(test_binary)
y_binary_test_predict = binary_clf.predict(X_binary_test)
print(classification_report(y_binary_test, y_binary_test_predict))
from sklearn.svm import SVC
from run_multilabel_classifier import run
param_grid = {
'bag_of_words__stop_words': ['english'],
'bag_of_words__ngram_range': [(1, 2)],
'bag_of_words__max_features': [500],
'dim_reduct__n_components': [300],
'normalizer__norm': ['l2'],
'classifier__C': [100., 0.1, 0.0001]
}
clf = SVC()
run(param_grid, clf)