인공신경망 파이썬 코드 공유

본 포스팅에선 인공신경망 실습 시 사용했던 파이썬 코드를 공유합니다.

import pandas as pd
from sklearn.neural_network import MLPClassifier
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler


data= pd.read_csv('C:\\Users\\Desktop\\파일명.csv', engine='python')

In [14]:

cols_to_keep=['churn','나이','성별','방문빈도','총구매금액','고객기간','결혼여부']
data1=data[cols_to_keep]
data1

Out[14]:

	churn	나이	성별	방문빈도	총구매금액	고객기간	결혼여부
0	1	57	1	1	237500	1158	0
1	1	50	1	1	610850	1206	0
2	1	39	1	1	19000	1880	0
3	1	49	1	1	9500	1348	0
4	1	42	0	3	589000	977	0
5	0	62	1	6	634600	908	0
6	1	46	1	1	779000	799	0
7	1	39	0	1	58000	1526	0
8	1	47	1	1	133000	967	0
9	1	68	0	2	137750	1936	0
10	0	48	1	1	89250	1386	0
11	1	53	0	2	1007000	1449	0
12	0	48	1	2	323200	1334	0
13	0	39	0	1	260400	1499	0
14	0	71	0	4	500750	1459	0
15	1	62	1	1	541500	999	0
16	1	53	1	2	1957000	1420	0
17	1	60	1	3	274500	1087	0
18	0	38	1	5	1932550	1494	0
19	0	40	0	1	63750	1522	0
20	0	30	1	1	9500	900	0
21	0	45	1	1	47500	1122	1
22	0	58	1	2	48000	1283	1
23	1	41	0	1	313500	1194	0
24	1	63	0	2	344800	2170	0
25	1	36	0	1	337250	1043	0
26	1	71	1	1	13500	1509	0
27	0	40	1	1	85500	1317	0
28	0	43	1	1	98800	913	0
29	0	28	0	1	184500	1101	0
…	…	…	…	…	…	…	…
4969	0	55	0	1	117000	-20	0
4970	0	25	1	1	468000	-21	0
4971	0	63	0	1	214200	-21	0
4972	0	38	1	1	202500	-18	0
4973	0	24	1	1	72000	-18	0
4974	0	59	1	1	222300	-27	0
4975	0	39	0	1	813000	-17	0
4976	0	34	1	1	89100	-19	0
4977	0	33	1	1	207000	-18	0
4978	0	36	0	1	117000	-24	0
4979	0	47	1	1	392000	-20	0
4980	0	33	0	2	103000	-16	0
4981	0	31	0	1	153000	-22	0
4982	0	29	1	1	54000	-17	0
4983	0	26	1	1	88000	-21	0
4984	0	49	0	1	207000	-22	0
4985	0	31	1	1	416000	-24	0
4986	0	45	0	1	304000	-17	0
4987	0	39	1	1	281000	-27	0
4988	0	76	0	1	40000	-24	0
4989	0	51	1	1	678000	582	0
4990	0	31	0	1	88000	-17	0
4991	0	40	0	2	477000	-17	0
4992	0	42	1	1	108000	-19	0
4993	0	32	1	1	475800	-23	0
4994	0	43	1	1	665000	-25	0
4995	0	32	0	1	225000	-22	0
4996	0	41	1	1	106200	-16	0
4997	0	27	1	1	306000	-16	0
4998	0	31	1	1	88000	-17	0

4999 rows × 7 columns

In [15]:

X=data1.drop(['churn'],axis=1)
Y=data1['churn']

X_train,X_test,Y_train,Y_test=train_test_split(X,Y,test_size=0.30)

In [16]:

sc=StandardScaler() #데이터 표준화
sc.fit(X_train) #트레이닝 데이터로 변환계수를 추정

Out[16]:

StandardScaler(copy=True, with_mean=True, with_std=True)

In [17]:

X_train_std=sc.transform(X_train) #standardscale.transform을 이용해 실제로 자료를 반환
X_test_std=sc.transform(X_test)

In [20]:

mlp = MLPClassifier() #다층퍼셉트론분류기
mlp.fit(X_train, Y_train)

print("학습용 데이터셋 정확도: {:.3f}".format(mlp.score(X_train, Y_train)))
print("검증용 데이터셋 정확도: {:.3f}".format(mlp.score(X_test, Y_test)))

학습용 데이터셋 정확도: 0.704
검증용 데이터셋 정확도: 0.713

In [21]:

data1["predict"] = mlp.predict(X)
data1

C:\Users\gksmf\anaconda3\lib\site-packages\ipykernel_launcher.py:1: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy
  """Entry point for launching an IPython kernel.

Out[21]:

	churn	나이	성별	방문빈도	총구매금액	고객기간	결혼여부	predict
0	1	57	1	1	237500	1158	0	1
1	1	50	1	1	610850	1206	0	0
2	1	39	1	1	19000	1880	0	1
3	1	49	1	1	9500	1348	0	1
4	1	42	0	3	589000	977	0	0
5	0	62	1	6	634600	908	0	0
6	1	46	1	1	779000	799	0	0
7	1	39	0	1	58000	1526	0	1
8	1	47	1	1	133000	967	0	1
9	1	68	0	2	137750	1936	0	1
10	0	48	1	1	89250	1386	0	1
11	1	53	0	2	1007000	1449	0	0
12	0	48	1	2	323200	1334	0	1
13	0	39	0	1	260400	1499	0	1
14	0	71	0	4	500750	1459	0	0
15	1	62	1	1	541500	999	0	0
16	1	53	1	2	1957000	1420	0	0
17	1	60	1	3	274500	1087	0	1
18	0	38	1	5	1932550	1494	0	0
19	0	40	0	1	63750	1522	0	1
20	0	30	1	1	9500	900	0	1
21	0	45	1	1	47500	1122	1	1
22	0	58	1	2	48000	1283	1	1
23	1	41	0	1	313500	1194	0	0
24	1	63	0	2	344800	2170	0	1
25	1	36	0	1	337250	1043	0	0
26	1	71	1	1	13500	1509	0	1
27	0	40	1	1	85500	1317	0	1
28	0	43	1	1	98800	913	0	1
29	0	28	0	1	184500	1101	0	1
…	…	…	…	…	…	…	…	…
4969	0	55	0	1	117000	-20	0	0
4970	0	25	1	1	468000	-21	0	0
4971	0	63	0	1	214200	-21	0	0
4972	0	38	1	1	202500	-18	0	0
4973	0	24	1	1	72000	-18	0	0
4974	0	59	1	1	222300	-27	0	0
4975	0	39	0	1	813000	-17	0	0
4976	0	34	1	1	89100	-19	0	0
4977	0	33	1	1	207000	-18	0	0
4978	0	36	0	1	117000	-24	0	0
4979	0	47	1	1	392000	-20	0	0
4980	0	33	0	2	103000	-16	0	0
4981	0	31	0	1	153000	-22	0	0
4982	0	29	1	1	54000	-17	0	0
4983	0	26	1	1	88000	-21	0	0
4984	0	49	0	1	207000	-22	0	0
4985	0	31	1	1	416000	-24	0	0
4986	0	45	0	1	304000	-17	0	0
4987	0	39	1	1	281000	-27	0	0
4988	0	76	0	1	40000	-24	0	0
4989	0	51	1	1	678000	582	0	0
4990	0	31	0	1	88000	-17	0	0
4991	0	40	0	2	477000	-17	0	0
4992	0	42	1	1	108000	-19	0	0
4993	0	32	1	1	475800	-23	0	0
4994	0	43	1	1	665000	-25	0	0
4995	0	32	0	1	225000	-22	0	0
4996	0	41	1	1	106200	-16	0	0
4997	0	27	1	1	306000	-16	0	0
4998	0	31	1	1	88000	-17	0	0

4999 rows × 8 columns