def GridSearch(self, model=modelMap.get('svm'), grid=None):
import time
st = time.time
import importlib
modelClassName = model.modelClass
modelClass = getattr(importlib.import_module(model.libPath),
modelClassName)
logger.info('Training the %s -- %s' % (modelClassName, model.type))
_classifier = modelClass(**model.params)
self.param_grid = {'C': [10, 100], 'gamma': [0.01, 0.1]}
self.grid = grid or GridSearchCV(
estimator=_classifier, param_grid=self.param_grid, cv=5)
self.grid.fit(self.X_train[..., self.adhocDependentColsLen:],
self.y_train)
with open(gridPath, 'wb') as saved_grid:
pickle.dump(self.grid,
saved_grid,
protocol=pickle.HIGHEST_PROTOCOL)
print("Test set score: {:.2f}".format(self.grid.score()))
print("best cross-validation score: {:.2f}".format(
self.grid.best_score_))
print("best parameter: {}".format(self.grid.best_params_))
print("best estimator: {}".format(self.grid.best_estimator_))
print("it took {} seconds".format(time.time - st))
def moreMetrics(self, model=modelMap.get('svm')):
global log
print("=" * 30)
print(model.modelClass)
print('****Results****')
acc = accuracy_score(self.y_test, self.result)
print("Accuracy: {:.4%}".format(acc))
ll = log_loss(self.y_test, self.result)
print("Log Loss: {}".format(ll))
log_entry = pd.DataFrame([[model.modelClass, acc * 100, ll]],
columns=log_cols)
log = log.append(log_entry)
print("=" * 30)
# horizontal bar plot
import seaborn as sns
sns.set_color_codes("muted")
sns.barplot(x='Accuracy', y='Classifier', data=log, color="b")
plt.xlabel('Accuracy %')
plt.title('Classifier Accuracy')
plt.show()
sns.set_color_codes("muted")
sns.barplot(x='Log Loss', y='Classifier', data=log, color="g")
plt.xlabel('Log Loss')
plt.title('Classifier Log Loss')
plt.show()
def getCrossValidationScore(self, model=modelMap.get('svm')):
from sklearn.model_selection import cross_val_score
import importlib
modelClassName = model.modelClass
modelClass = getattr(importlib.import_module(model.libPath),
modelClassName)
logger.info('Training the %s -- %s' % (modelClassName, model.type))
_classifier = modelClass(**model.params)
scores = cross_val_score(_classifier,
self.X_train[...,
self.adhocDependentColsLen:],
self.y_train,
cv=5)
print("Mean cross-validation accuracy: {:.2f}".format(np.mean(scores)))
def modelTraining(self, model=modelMap.get('svm'), saved=None):
import importlib
modelClassName = model.modelClass
modelClass = getattr(importlib.import_module(model.libPath),
modelClassName)
logger.info('Training the %s -- %s' % (modelClassName, model.type))
self.classifier = modelClass(**model.params)
self.classifier = saved or self.classifier.fit(
self.X_train[..., self.adhocDependentColsLen:], self.y_train)
logger.info('Creating a pickle of trained classifier')
with open(model.picklePath, 'wb') as saved_predictor:
pickle.dump(self.classifier,
saved_predictor,
protocol=pickle.HIGHEST_PROTOCOL)
def getSavedPredictor(predictor='svm'):
model = modelMap.get(predictor)
with open(model.picklePath, 'rb') as handle:
saved_predictor = pickle.load(handle)
return saved_predictor