은행의 전화 마케팅에 대해 고객의 반응 여부

data

data

display(train.head())
display(test.head())
display(submission.head())

train

	ID	age	job	marital	education	default	balance	housing	loan	contact	day	month	campaign	pdays	previous	poutcome	y
0	13829	29	technician	single	tertiary	no	18254	no	no	cellular	11	may	2	-1	0	unknown	no
1	22677	26	services	single	secondary	no	512	yes	yes	unknown	5	jun	3	-1	0	unknown	no
2	10541	30	management	single	secondary	no	135	no	no	cellular	14	aug	2	-1	0	unknown	no
3	13689	41	technician	married	unknown	no	30	yes	no	cellular	10	jul	1	-1	0	unknown	no
4	11304	27	admin.	single	secondary	no	321	no	yes	unknown	2	sep	1	-1	0	unknown	no

test

	ID	age	job	marital	education	default	balance	housing	loan	contact	day	month	campaign	pdays	previous	poutcome
0	53608	32	management	single	tertiary	no	12569	no	no	cellular	1	jul	2	295	2	success
1	51055	25	services	single	secondary	no	801	no	no	cellular	5	jun	2	-1	0	unknown
2	52573	46	blue-collar	married	secondary	no	1728	yes	no	unknown	26	may	2	-1	0	unknown
3	50458	39	management	divorced	secondary	no	51	no	no	unknown	17	jun	2	-1	0	unknown
4	52272	31	services	single	tertiary	no	1626	no	no	unknown	31	jul	1	-1	0	unknown

submission

	ID
0	53608
1	51055
2	52573
3	50458
4	52272

데이터 분리

from sklearn.model_selection import train_test_split

x = train.drop(columns = ['ID', 'y'])
# drop(['ID', 'y'], axis = 1)
xd = pd.get_dummies(x)
y = train['y']

X_train, X_test, y_train, y_test = train_test_split(xd, y, stratify = y, random_state = 1)

RandomForest

predict_proba(x_test)

from sklearn.ensemble import RandomForestClassifier
model = RandomForestClassifier()
model.fit(X_train, y_train)
pred = model.predict_proba(x_test) 

평가

from sklearn.metrics import roc_auc_score, classification_report

print('test roc score : ',roc_auc_score(y_test,pred[:,1]))

## +
print(f'test roc score : {roc_auc_score(y_test, pred[:,1]):.3f}')
print(f'test matrix report : \n {classification_report(y_test, model.predict(X_test))}')

# result
test roc score :  0.7756576420890937

test

pred_test = model.predict_proba(pd.get_dummies(test.drop('ID', axis = 1)))
submission['predict'] = pred_test[:,1]

submission

display(submission.head())
submission.to_csv('submission_.csv', index=False)

	ID	predict
0	53608	0.73
1	51055	0.82
2	52573	0.00
3	50458	0.14
4	52272	0.32