def run_knn(df_knn):
print(
"\n\n----------------------K Nearest Neighbors----------------------\n\n"
)
x = np.array(df_knn.ix[:, 0:])
y = np.array(df_knn['class'])
x_train, x_test, y_train, y_test = split_train_test(x, y)
neighbors = [
1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31, 33, 35, 37,
39, 41, 43, 45, 47, 49
]
cv_scores = []
for k in neighbors:
knn = KNeighborsClassifier(n_neighbors=k)
scores = cross_val_score(knn,
x_train,
y_train,
cv=10,
scoring='accuracy')
cv_scores.append(scores.mean())
MSE = [1 - x for x in cv_scores]
optimal_k = neighbors[MSE.index(min(MSE))]
print("The optimal number of neighbors is %d" % optimal_k)
plot_data(neighbors, MSE)
def run_random_forest(df_knn):
print("\n\n----------------------Random Forest----------------------\n\n")
x = np.array(df_knn.ix[:, 0:])
y = np.array(df_knn['class'])
x_train, x_test, y_train, y_test = split_train_test(x,y)
rf = RandomForestClassifier(random_state=1, n_estimators=250, min_samples_split=8, min_samples_leaf=4)
rf.fit(x_train,y_train)
pred = rf.predict(x_test)
print("Accuracy: ",accuracy_score(y_test, pred))
def run_random_forest(df_knn):
print("\n\n----------------------Random Forest----------------------\n\n")
y = np.array(df_knn["class"])
x = np.array(df_knn.drop(columns="class"))
x_train, x_test, y_train, y_test = split_train_test(x, y)
rf = RandomForestClassifier(random_state=1,
n_estimators=250,
min_samples_split=8,
min_samples_leaf=4)
rf.fit(x_train, y_train)
pred = rf.predict(x_test)
test_pred = rf.predict(x_train)
success_indices = np.where(y_train == 2)[0]
test_frame = x_train[success_indices]
print("location index:", success_indices)
tst_y = y_train[success_indices]
pred_frame = rf.predict(test_frame)
print("\n\n----------------------Pred success----------------------\n\n",
accuracy_score(tst_y, pred_frame))
failure_indices = np.where(y_train == 1)[0]
test_frame = x_train[failure_indices]
tst_y = y_train[failure_indices]
pred_frame = rf.predict(test_frame)
print("\n\n----------------------Pred Failure----------------------\n\n",
accuracy_score(tst_y, pred_frame))
good_indices = np.where(y_train == 1)[0]
test_frame = x_train[good_indices]
tst_y = y_train[good_indices]
pred_frame = rf.predict(test_frame)
print("\n\n----------------------Pred good----------------------\n\n",
accuracy_score(tst_y, pred_frame))
# print the Training Accuracy
print(
"\n\n----------------------Training Set Accuracy----------------------\n\n",
accuracy_score(y_train, test_pred),
)
print(
"\n\n----------------------Testing Set Accuracy----------------------\n\n",
accuracy_score(y_test, pred),
)
return rf
def run_logistic_regression(df_knn):
y = np.array(df_knn["class"])
x = np.array(df_knn.drop(columns="class"))
x_train, x_test, y_train, y_test = split_train_test(x, y)
logistic = LogisticRegression()
logistic.fit(x_train, y_train)
pred_train = logistic.predict(x_train)
print ("Training accuracy: ", (logistic.score(x_train, y_train) * 100))
pred_test = logistic.predict(x_test)
print ("Testing accuracy: ", (logistic.score(x_test, y_test) * 100))
success_indices = np.where(y_train==2)[0]
test_frame = x_train[success_indices]
tst_y = y_train[success_indices]
pred_frame = logistic.predict(test_frame)
print(
"\n\n----------------------Pred success----------------------\n\n",
logistic.score(test_frame, tst_y)
)
failure_indices = np.where(y_train==1)[0]
test_frame = x_train[failure_indices]
tst_y = y_train[failure_indices]
pred_frame = logistic.predict(test_frame)
print(
"\n\n----------------------Pred Failure----------------------\n\n",
logistic.score(test_frame, tst_y)
)
good_indices = np.where(y_train==1)[0]
test_frame = x_train[good_indices]
tst_y = y_train[good_indices]
pred_frame = logistic.predict(test_frame)
print(
"\n\n----------------------Pred good----------------------\n\n",
logistic.score(test_frame, tst_y)
)
return logistic
def run_xgboost_imdb(df_knn):
print(
"\n\n----------------------XGBoost on IMDB dataset----------------------\n\n"
)
x = np.array(df_knn.ix[:, 0:])
y = np.array(df_knn['class'])
x_train, x_test, y_train, y_test = split_train_test(x, y)
x_train = np.delete(x_train, [0, 1, 2, 3, 4, 5, 9, 44], axis=1)
x_test = np.delete(x_test, [0, 1, 2, 3, 4, 5, 9, 44], axis=1)
model = xgboost.XGBClassifier()
model.fit(x_train, y_train)
pred = model.predict(x_train)
accuracy = accuracy_score(y_train, pred)
print("Training accuracy: %.2f%%" % (accuracy * 100.0))
pred = model.predict(x_test)
accuracy = accuracy_score(y_test, pred)
print("Testing accuracy: %.2f%%" % (accuracy * 100.0))
# !/usr/bin/python
# _*_ coding:utf-8 _*_
import numpy
print numpy.version.version
from split_dataset import split_train_test
split_train_test('file1.txt')