def train_random_forest(X, y):
from sklearn.ensemble import RandomForestClassifier
from sklearn.grid_search import GridSearchCV
from sklearn.cross_validation import train_test_split
# train - test split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.10)
# TODO: separate feature selection step
# fit random forest classifier
rfc = RandomForestClassifier(n_estimators=2000, n_jobs=-1, oob_score=True)
param_grid = {
'max_features': ['auto', 'sqrt', 'log2'],
'min_samples_leaf': [20, 35, 50]
}
CV_rfc = GridSearchCV(estimator=rfc, param_grid=param_grid, cv=5)
CV_rfc.fit(X_train, y_train)
print('Random Forest classifier best parameters:', CV_rfc.best_params_)
# normalized discounted cumulative gain
predictions = format_predictions(X_test, CV_rfc)
ndcg = ndcg_n(predictions, y_test)
print('Random Forest classifier NDCG (on test set): {0:.2f}'.format(ndcg))
return CV_rfc
def train_logistic_regression(X, y):
from sklearn.linear_model import LogisticRegression
from sklearn.cross_validation import train_test_split
# train - test split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.10)
# TODO: separate feature selection step
# fit logistic regression
lr = LogisticRegression(multi_class='multinomial', solver='lbfgs', penalty='l2')
lr.fit(X_train, y_train)
# normalized discounted cumulative gain
predictions = format_predictions(X_test, lr)
ndcg = ndcg_n(predictions, y_test)
print('Logistic regression classifier NDCG (on test set): : {0:.4f}'.format(ndcg))
return lr
def train_decision_tree(X, y):
from sklearn import tree
from sklearn.grid_search import GridSearchCV
from sklearn.cross_validation import train_test_split
# train - test split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.10)
# TODO: separate feature selection step
# fit decision tree
tree = tree.DecisionTreeClassifier()
param_grid = {'max_depth': [3, 5, 7, 9]}
CV_tree = GridSearchCV(estimator=tree, param_grid=param_grid, cv=5)
CV_tree.fit(X_train, y_train)
print('Decision tree classifier best parameters:', CV_tree.best_params_)
# normalized discounted cumulative gain
predictions = format_predictions(X_test, CV_tree)
ndcg = ndcg_n(predictions, y_test)
print('Decision tree classifier NDCG (on test set): {0:.2f}'.format(ndcg))
return CV_tree