from knock52 import read_data
from sklearn import metrics
import joblib
def accuracy(feature, label, model_name, vectorizer_name):
model = joblib.load(model_name)
vectorizer = joblib.load(vectorizer_name)
x = vectorizer.transform(feature)
prediction = model.predict(x)
return metrics.accuracy_score(label, prediction)
if __name__ == "__main__":
train = open('train.feature.txt')
train_ftr, train_label = read_data(train)
print('Train Accuracy: ' + str(
round(accuracy(train_ftr, train_label, 'model.pkl', 'vectorizer.pkl'),
6)))
test = open('test.feature.txt')
test_ftr, test_label = read_data(test)
print('Test Accuracy: ' + str(
round(accuracy(test_ftr, test_label, 'model.pkl', 'vectorizer.pkl'),
6)))
'''
Train Accuracy: 0.996441
Test Accuracy: 0.828571
'''
from knock52 import read_data
import joblib
def test(feature):
model = joblib.load('model.pkl')
print('Prediction: ' + str(model.predict(feature)))
print('Prediction Probability: ' + str(model.predict_proba(feature)))
if __name__ == "__main__":
text = open('test.feature.txt')
feature, label = read_data(text)
vectorizer = joblib.load('vectorizer.pkl')
x_test = vectorizer.transform(feature)
test(x_test)
from knock52 import read_data
import joblib
from sklearn import metrics
def calc_score(feature, label):
model = joblib.load('model.pkl')
vectorizer = joblib.load('vectorizer.pkl')
x = vectorizer.transform(feature)
prediction = model.predict(x)
print(metrics.classification_report(label, prediction))
if __name__ == "__main__":
test = open('test.feature.txt')
test_ftr, test_label = read_data(test)
calc_score(test_ftr, test_label)
from knock52 import read_data, vectorize
import joblib
from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier
from sklearn.svm import LinearSVC
from sklearn.naive_bayes import MultinomialNB
# Search for the training algorithms and
# parameters that achieves the best accuracy score on the validation data.
# Then compute its accuracy score on the test data.
if __name__ == "__main__":
vectorizer = joblib.load('vectorizer.pkl')
train = open('train.feature.txt')
train_ftr, train_label = read_data(train)
valid = open('valid.feature.txt')
valid_ftr, valid_label = read_data(valid)
test = open('test.feature.txt')
test_ftr, test_label = read_data(test)
x_train = vectorizer.transform(train_ftr)
y_train = train_label
x_valid = vectorizer.transform(valid_ftr)
y_valid = valid_label
x_test = vectorizer.transform(test_ftr)
y_test = test_label
# Multinomial NB
nb_model = MultinomialNB()
nb_model.fit(x_train, y_train)