도찐개찐

[머신러닝] 05. 데이터 전처리 본문

PYTHON/데이터분석

[머신러닝] 05. 데이터 전처리

도개진 2023. 1. 3. 09:37 
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import sklearn

데이터전처리

  • 머신러닝 알고리즘을 익히는 것 못치 않게
    데이터 전처리 역시 중요한 과정 중에 하나
  • 무엇보다 머신러닝 알고리즘을 적용하기 전에
    데이터에 대해 미리 처리해야 하는 기본사항이 존재
  • 결측치 처리
    • NaN, Null은 허용되지 않음
  • 원핫인코딩
    • 머신러닝 알고리즘들은 문자열값을 데이터의 입력값으로 허용하지 않음
    • 따라서, 모든 문자열값은 인코딩해서 숫자형으로 변환해둬야 함
    • 한편, 텍스트 데이터들은 벡터화해서 처리
    • 머신러닝을 위한 인코딩은 레이블인코딩 과 원핫인코딩 등이 있음

레이블 인코딩

  • 범주형값을 숫자형값으로 변환함
from sklearn.preprocessing import LabelEncoder
items = ['티비','냉장고','가스렌지','에어콘','컴퓨터']
encoder = LabelEncoder()
encoder.fit(items)
lables = encoder.transform(items)
lables
array([4, 1, 0, 2, 3])
# 인코딩 순서 확인
encoder.classes_
array(['가스렌지', '냉장고', '에어콘', '컴퓨터', '티비'], dtype='<U4')

LabelEncoder 시용시 문제점

  • 문자열값을 숫자형으로 변환시켰을때 발생할 수 있는 문제는
    각 값의 대소 관계를 통해 중요도 여부가 존재할 수 있음
  • 즉, 인코딩 된 값에 서수척도가 생길수 있음
  • 따라서, 대소관계가 있는 데이터를 분석할 경우 정확도에 영향을 미칠수있음
    • => 원핫인코딩을 사용함으로써 문제해결
'티비','냉장고','가스렌지','에어콘','컴퓨터'   
1    0    0    0    0   
0    1    0    0    0   
0    0    1    0    0   
0    0    0    1    0   
0    0    0    0    1  
from sklearn.preprocessing import OneHotEncoder
# 먼저, LabelEncoder로 문자열을 숫자값으로 변환해두어야 함!
# 그런 다음, 1차원 데이터를 2차원 데이터 변환
lables = lables.reshape(-1, 1)   # -1 : 행을 열로 변환 (가로->세로)
lables
array([[4],
       [1],
       [0],
       [2],
       [3]])
onehot = OneHotEncoder()
onehot.fit(lables)
onlables = onehot.transform(lables)

array([[0., 0., 0., 0., 1.],
       [0., 1., 0., 0., 0.],
       [1., 0., 0., 0., 0.],
       [0., 0., 1., 0., 0.],
       [0., 0., 0., 1., 0.]])
# ['티비','냉장고','가스렌지','에어콘','컴퓨터']
onlables.toarray()
array([[0., 0., 0., 0., 1.],
       [0., 1., 0., 0., 0.],
       [1., 0., 0., 0., 0.],
       [0., 0., 1., 0., 0.],
       [0., 0., 0., 1., 0.]])

pandas의 원핫인코딩

  • get_dummies 함수 사용
  • 단, 변환대상은 반드시 데이터프레임이어야 함!
df = pd.DataFrame({'': items})
pd.get_dummies(df)
  _가스렌지 _냉장고 _에어콘 _컴퓨터 _티비
0 0 0 0 0 1
1 0 1 0 0 0
2 1 0 0 0 0
3 0 0 1 0 0
4 0 0 0 1 0

특성 스케일링과 표준화/정규화

  • 서로 다른 범위, 단위의 변수값을
    일정수준으로 맞추는 작업을 특성feature 스케일링이라 함
  • 어떤 데이터의 값이 정수와 실수가 혼용되어 있거나
  • 값의 범위가 1 ~ 100, 0 ~ 0.001, 1 ~ 10000 등등의 경우
  • 데이터 분석시 많은 cpu 파워/메모리가 필요하고
  • 학습시 느려질수 있으며 제대로 된 결과가 나오지 않을 수 있음
  • 이것을 제대로 변환하는 방법은 정규화 와 표준화 가 있음

표준화

  • 특성값을 평균이 0이고 표준편차가 1인 정규분포를 가진값으로 변환하는 것
X = np.arange(9) - 3
X
array([-3, -2, -1,  0,  1,  2,  3,  4,  5])
np.mean(X, axis=0), np.std(X, axis=0)
(1.0, 2.581988897471611)
from sklearn.preprocessing import StandardScaler
X = X.reshape(-1, 1)
scaler = StandardScaler()
scaler.fit(X) # wide to long 변환
XX = scaler.transform(X)
XX
array([[-1.54919334],
       [-1.161895  ],
       [-0.77459667],
       [-0.38729833],
       [ 0.        ],
       [ 0.38729833],
       [ 0.77459667],
       [ 1.161895  ],
       [ 1.54919334]])
print(np.min(XX, axis=0))
print(np.max(XX, axis=0))
[-1.54919334]
[1.54919334]

정규화

  • 특성값을 최소 0, 최대 1사이의 값으로 변환하는 것
  • 데이터의 분포가 정규분포를 따르지 않을때 적용
from sklearn.preprocessing import MinMaxScaler
X = X.reshape(-1, 1)
scaler = MinMaxScaler()
scaler.fit(X) # wide to long 변환
XXX = scaler.transform(X)
print(np.min(XXX, axis=0))
print(np.max(XXX, axis=0))
[0.]
[1.]

중앙값, 사분위수 기반 스케일링

  • RobustScaler
  • 이상치나 극단치에 영향을 덜 받는 스케일링 기법
from sklearn.preprocessing import RobustScaler
X = X.reshape(-1, 1)
scaler = RobustScaler()
scaler.fit(X) # wide to long 변환
XXXX = scaler.transform(X)
print(np.min(XXXX, axis=0))
print(np.max(XXXX, axis=0))
[-1.]
[1.]
from sklearn.preprocessing import LabelEncoder

from sklearn.model_selection import train_test_split
from sklearn.model_selection import cross_val_score

from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier

from sklearn.metrics import accuracy_score
titanic = pd.read_csv('../data/titanic2.csv')

표준화 적용

titanic.head()
data = titanic.iloc[:, [1, 5,6, 7, 8, 11, 12, 13]]
scaler = StandardScaler()
scaler.fit(data)
data2 = scaler.transform(data)
target = titanic.survived
X_train, X_test, y_train, y_test = train_test_split(data2, target, train_size = 0.7,
                stratify=target)

dtclf = DecisionTreeClassifier()
dtclf.fit(X_train, y_train)
pred = dtclf.predict(X_test)

accuracy_score(y_test, pred)
0.7551020408163265
rfclf = RandomForestClassifier()
rfclf.fit(X_train, y_train)
pred = rfclf.predict(X_test)

accuracy_score(y_test, pred)
0.7882653061224489
rfclf = RandomForestClassifier()
scores = cross_val_score(rfclf, data2, target, scoring='accuracy', cv = 10)
np.mean(scores)
0.7425249559600705

정규화 적용

data = titanic.iloc[:, [1, 5,6, 7, 8, 11, 12, 13]]
scaler = MinMaxScaler()
scaler.fit(data)
data2 = scaler.transform(data)
target = titanic.survived
X_train, X_test, y_train, y_test = train_test_split(data2, target, train_size = 0.7,
                stratify=target)

dtclf = DecisionTreeClassifier()
dtclf.fit(X_train, y_train)
pred = dtclf.predict(X_test)

accuracy_score(y_test, pred)
0.7729591836734694
rfclf = RandomForestClassifier()
rfclf.fit(X_train, y_train)
pred = rfclf.predict(X_test)

accuracy_score(y_test, pred)
0.7908163265306123
rfclf = RandomForestClassifier()
scores = cross_val_score(rfclf, data2, target, scoring='accuracy', cv = 10)
np.mean(scores)
0.7448502642395772

중앙값 스케일러

data = titanic.iloc[:, [1, 5,6, 7, 8, 11, 12, 13]]
scaler = RobustScaler()
scaler.fit(data)
data2 = scaler.transform(data)
target = titanic.survived
X_train, X_test, y_train, y_test = train_test_split(data2, target, train_size = 0.7,
                stratify=target)

dtclf = DecisionTreeClassifier()
dtclf.fit(X_train, y_train)
pred = dtclf.predict(X_test)

accuracy_score(y_test, pred)
0.7780612244897959
rfclf = RandomForestClassifier()
rfclf.fit(X_train, y_train)
pred = rfclf.predict(X_test)

accuracy_score(y_test, pred)
0.7908163265306123
rfclf = RandomForestClassifier()
scores = cross_val_score(rfclf, data2, target, scoring='accuracy', cv = 10)
np.mean(scores)
0.7364004697592483
titanic.head()
t_age = np.array(titanic.age)
t_age = t_age.reshape(-1, 1)
t_fare = np.array(titanic.fare)
t_fare = t_fare.reshape(-1, 1)

scaler = StandardScaler()
scaler.fit(t_age) # wide to long 변환
scaler.fit(t_fare)
s_age = scaler.transform(t_age)
s_fare = scaler.transform(t_fare)
titanic['age_scale'] = s_age
titanic['fare_scale'] = s_fare

titanic
data = titanic.iloc[:, [1, 6, 7, 11, 12, 13, 14, 15]]
target = titanic.survived
data
  pclass sibsp parch gender Embarked Title age_scale fare_scale
0 1 0 0 0 2 10 -0.081628 3.442063
1 1 1 2 1 2 9 -0.624341 2.286663
2 1 1 2 0 2 10 -0.603406 2.286663
3 1 1 2 1 2 13 -0.062303 2.286663
4 1 1 2 0 2 14 -0.158929 2.286663
... ... ... ... ... ... ... ... ...
1301 3 1 0 0 0 10 -0.361842 -0.362727
1302 3 1 0 0 0 10 -0.100953 -0.362727
1303 3 0 0 1 0 13 -0.129941 -0.502433
1304 3 0 0 1 0 13 -0.120279 -0.502433
1305 3 0 0 1 2 13 -0.081628 -0.489871

1306 rows × 8 columns

from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(data, target, train_size = 0.7,
                stratify=target, random_state=2211161315)
from sklearn.tree import DecisionTreeClassifier
dtclf = DecisionTreeClassifier(random_state=2211161315)
dtclf.fit(X_train, y_train)
pred = dtclf.predict(X_test)

accuracy_score(y_test, pred)
0.7448979591836735
scaler = MinMaxScaler()
scaler.fit(t_age) # wide to long 변환
scaler.fit(t_fare)
s_age = scaler.transform(t_age)
s_fare = scaler.transform(t_fare)
titanic['age_scale'] = s_age
titanic['fare_scale'] = s_fare

titanic
data = titanic.iloc[:, [1, 6, 7, 11, 12, 13, 14, 15]]
target = titanic.survived
  Unnamed: 0 pclass survived name sex age sibsp parch fare embarked title gender Embarked Title age_scale fare_scale
0 0 1 1 Allen, Miss. Elisabeth Walton female 29.0000 0 0 211.3375 S Miss 0 2 10 0.056604 0.412503
1 1 1 1 Allison, Master. Hudson Trevor male 0.9167 1 2 151.5500 S Master 1 2 9 0.001789 0.295806
2 2 1 0 Allison, Miss. Helen Loraine female 2.0000 1 2 151.5500 S Miss 0 2 10 0.003904 0.295806
3 3 1 0 Allison, Mr. Hudson Joshua Creighton male 30.0000 1 2 151.5500 S Mr 1 2 13 0.058556 0.295806
4 4 1 0 Allison, Mrs. Hudson J C (Bessie Waldo Daniels) female 25.0000 1 2 151.5500 S Mrs 0 2 14 0.048797 0.295806
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
1301 1304 3 0 Zabour, Miss. Hileni female 14.5000 1 0 14.4542 C Miss 0 0 10 0.028302 0.028213
1302 1305 3 0 Zabour, Miss. Thamine female 28.0000 1 0 14.4542 C Miss 0 0 10 0.054652 0.028213
1303 1306 3 0 Zakarian, Mr. Mapriededer male 26.5000 0 0 7.2250 C Mr 1 0 13 0.051725 0.014102
1304 1307 3 0 Zakarian, Mr. Ortin male 27.0000 0 0 7.2250 C Mr 1 0 13 0.052700 0.014102
1305 1308 3 0 Zimmerman, Mr. Leo male 29.0000 0 0 7.8750 S Mr 1 2 13 0.056604 0.015371

1306 rows × 16 columns

X_train, X_test, y_train, y_test = train_test_split(data, target, train_size = 0.7,
                stratify=target, random_state=2211161315)
scaler = RobustScaler()
scaler.fit(t_age) # wide to long 변환
XX = scaler.transform(t_age)
titanic['age_scale'] = XX
titanic
  Unnamed: 0 pclass survived name sex age sibsp parch fare embarked title gender Embarked Title age_scale
0 0 1 1 Allen, Miss. Elisabeth Walton female 29.0000 0 0 211.3375 S Miss 0 2 10 0.076923
1 1 1 1 Allison, Master. Hudson Trevor male 0.9167 1 2 151.5500 S Master 1 2 9 -2.083331
2 2 1 0 Allison, Miss. Helen Loraine female 2.0000 1 2 151.5500 S Miss 0 2 10 -2.000000
3 3 1 0 Allison, Mr. Hudson Joshua Creighton male 30.0000 1 2 151.5500 S Mr 1 2 13 0.153846
4 4 1 0 Allison, Mrs. Hudson J C (Bessie Waldo Daniels) female 25.0000 1 2 151.5500 S Mrs 0 2 14 -0.230769
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
1301 1304 3 0 Zabour, Miss. Hileni female 14.5000 1 0 14.4542 C Miss 0 0 10 -1.038462
1302 1305 3 0 Zabour, Miss. Thamine female 28.0000 1 0 14.4542 C Miss 0 0 10 0.000000
1303 1306 3 0 Zakarian, Mr. Mapriededer male 26.5000 0 0 7.2250 C Mr 1 0 13 -0.115385
1304 1307 3 0 Zakarian, Mr. Ortin male 27.0000 0 0 7.2250 C Mr 1 0 13 -0.076923
1305 1308 3 0 Zimmerman, Mr. Leo male 29.0000 0 0 7.8750 S Mr 1 2 13 0.076923

1306 rows × 15 columns

승객의 나이, 요금을 범주형으로 변환

# 나이범주 : 5, 10, 20, 30, 40, 50, 60, 70, 80, 90~
def getAge(x):
    result = 'a'
    if 80 < x <= 90:
        result = 'b'
    elif 70 < x <= 80:
        result = 'c'
    elif 60 < x <= 70:
        result = 'd'
    elif 50 < x <= 60:
        result = 'e'
    elif 40 < x <= 50:
        result = 'f'
    elif 30 < x <= 40:
        result = 'g'
    elif 20 < x <= 30:
        result = 'h'
    elif 10 < x <= 20:
        result = 'i'
    elif 5 < x <= 10:
        result = 'j'
    elif x <= 5:
        result = 'k'
    return result
titanic['Age_Area'] = titanic.age.apply(lambda x: getAge(x))
titanic
  Unnamed: 0 pclass survived name sex age sibsp parch fare embarked title gender Embarked Title Age_Area Fare_Area
0 0 1 1 Allen, Miss. Elisabeth Walton female 29.0000 0 0 211.3375 S Miss 0 2 10 h 4
1 1 1 1 Allison, Master. Hudson Trevor male 0.9167 1 2 151.5500 S Master 1 2 9 k 5
2 2 1 0 Allison, Miss. Helen Loraine female 2.0000 1 2 151.5500 S Miss 0 2 10 k 5
3 3 1 0 Allison, Mr. Hudson Joshua Creighton male 30.0000 1 2 151.5500 S Mr 1 2 13 h 5
4 4 1 0 Allison, Mrs. Hudson J C (Bessie Waldo Daniels) female 25.0000 1 2 151.5500 S Mrs 0 2 14 h 5
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
1301 1304 3 0 Zabour, Miss. Hileni female 14.5000 1 0 14.4542 C Miss 0 0 10 i 6
1302 1305 3 0 Zabour, Miss. Thamine female 28.0000 1 0 14.4542 C Miss 0 0 10 h 6
1303 1306 3 0 Zakarian, Mr. Mapriededer male 26.5000 0 0 7.2250 C Mr 1 0 13 h 6
1304 1307 3 0 Zakarian, Mr. Ortin male 27.0000 0 0 7.2250 C Mr 1 0 13 h 6
1305 1308 3 0 Zimmerman, Mr. Leo male 29.0000 0 0 7.8750 S Mr 1 2 13 h 6

1306 rows × 16 columns

# 요금범주 : 100, 200, 300, 400, 500
def getFare(x):
    result = 'a'
    if 400 < x <= 500:
        result = 'b'
    elif 300 < x <= 400:
        result = 'c'
    elif 200 < x <= 300:
        result = 'd'
    elif 100 < x <= 200:
        result = 'e'
    elif x <= 100:
        result = 'f'
    return result
titanic['Fare_Area'] = titanic.fare.apply(lambda x: getFare(x))
titanic
  Unnamed: 0 pclass survived name sex age sibsp parch fare embarked title gender Embarked Title Age_Area Fare_Area
0 0 1 1 Allen, Miss. Elisabeth Walton female 29.0000 0 0 211.3375 S Miss 0 2 10 h d
1 1 1 1 Allison, Master. Hudson Trevor male 0.9167 1 2 151.5500 S Master 1 2 9 k e
2 2 1 0 Allison, Miss. Helen Loraine female 2.0000 1 2 151.5500 S Miss 0 2 10 k e
3 3 1 0 Allison, Mr. Hudson Joshua Creighton male 30.0000 1 2 151.5500 S Mr 1 2 13 h e
4 4 1 0 Allison, Mrs. Hudson J C (Bessie Waldo Daniels) female 25.0000 1 2 151.5500 S Mrs 0 2 14 h e
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
1301 1304 3 0 Zabour, Miss. Hileni female 14.5000 1 0 14.4542 C Miss 0 0 10 i f
1302 1305 3 0 Zabour, Miss. Thamine female 28.0000 1 0 14.4542 C Miss 0 0 10 h f
1303 1306 3 0 Zakarian, Mr. Mapriededer male 26.5000 0 0 7.2250 C Mr 1 0 13 h f
1304 1307 3 0 Zakarian, Mr. Ortin male 27.0000 0 0 7.2250 C Mr 1 0 13 h f
1305 1308 3 0 Zimmerman, Mr. Leo male 29.0000 0 0 7.8750 S Mr 1 2 13 h f

1306 rows × 16 columns

encoder = LabelEncoder()
encoder.fit(titanic.Age_Area)
titanic['Age_Area'] = encoder.transform(titanic.Age_Area)
encoder = LabelEncoder()
encoder.fit(titanic.Fare_Area)
titanic['Fare_Area'] = encoder.transform(titanic.Fare_Area)
titanic.head()
  Unnamed: 0 pclass survived name sex age sibsp parch fare embarked title gender Embarked Title Age_Area Fare_Area
0 0 1 1 Allen, Miss. Elisabeth Walton female 29.0000 0 0 211.3375 S Miss 0 2 10 5 1
1 1 1 1 Allison, Master. Hudson Trevor male 0.9167 1 2 151.5500 S Master 1 2 9 8 2
2 2 1 0 Allison, Miss. Helen Loraine female 2.0000 1 2 151.5500 S Miss 0 2 10 8 2
3 3 1 0 Allison, Mr. Hudson Joshua Creighton male 30.0000 1 2 151.5500 S Mr 1 2 13 5 2
4 4 1 0 Allison, Mrs. Hudson J C (Bessie Waldo Daniels) female 25.0000 1 2 151.5500 S Mrs 0 2 14 5 2 
data = titanic.iloc[:, [1, 6, 7, 11, 12, 13, 14, 15]]
# data
scaler = StandardScaler()
scaler.fit(data)
data2 = scaler.transform(data)
target = titanic.survived
pd.DataFrame(data2)
  0 1 2 3 4 5 6 7
0 -1.549301 -0.479763 -0.445612 -1.347091 0.623271 -1.203895 0.220416 -4.794691
1 -1.549301 0.479763 1.863789 0.742340 0.623271 -1.738323 2.384797 -2.275839
2 -1.549301 0.479763 1.863789 -1.347091 0.623271 -1.203895 2.384797 -2.275839
3 -1.549301 0.479763 1.863789 0.742340 0.623271 0.399389 0.220416 -2.275839
4 -1.549301 0.479763 1.863789 -1.347091 0.623271 0.933817 0.220416 -2.275839
... ... ... ... ... ... ... ... ...
1301 0.840997 0.479763 -0.445612 -1.347091 -1.832210 -1.203895 0.941876 0.243013
1302 0.840997 0.479763 -0.445612 -1.347091 -1.832210 -1.203895 0.220416 0.243013
1303 0.840997 -0.479763 -0.445612 0.742340 -1.832210 0.399389 0.220416 0.243013
1304 0.840997 -0.479763 -0.445612 0.742340 -1.832210 0.399389 0.220416 0.243013
1305 0.840997 -0.479763 -0.445612 0.742340 0.623271 0.399389 0.220416 0.243013

1306 rows × 8 columns

X_train, X_test, y_train, y_test = train_test_split(data2, target, train_size = 0.7,
                stratify=target)

dtclf = RandomForestClassifier()
dtclf.fit(X_train, y_train)
pred = dtclf.predict(X_test)

accuracy_score(y_test, pred)
0.8061224489795918
data = titanic.iloc[:, [1, 6, 7, 11, 12, 13, 14, 15]]
# data
scaler = MinMaxScaler()
scaler.fit(data)
data2 = scaler.transform(data)
target = titanic.survived
pd.DataFrame(data2)
  0 1 2 3 4 5 6 7
0 0.0 0.000 0.000000 0.0 1.0 0.588235 0.625 0.333333
1 0.0 0.125 0.222222 1.0 1.0 0.529412 1.000 0.666667
2 0.0 0.125 0.222222 0.0 1.0 0.588235 1.000 0.666667
3 0.0 0.125 0.222222 1.0 1.0 0.764706 0.625 0.666667
4 0.0 0.125 0.222222 0.0 1.0 0.823529 0.625 0.666667
... ... ... ... ... ... ... ... ...
1301 1.0 0.125 0.000000 0.0 0.0 0.588235 0.750 1.000000
1302 1.0 0.125 0.000000 0.0 0.0 0.588235 0.625 1.000000
1303 1.0 0.000 0.000000 1.0 0.0 0.764706 0.625 1.000000
1304 1.0 0.000 0.000000 1.0 0.0 0.764706 0.625 1.000000
1305 1.0 0.000 0.000000 1.0 1.0 0.764706 0.625 1.000000

1306 rows × 8 columns

X_train, X_test, y_train, y_test = train_test_split(data2, target, train_size = 0.7,
                stratify=target)

dtclf = RandomForestClassifier()
dtclf.fit(X_train, y_train)
pred = dtclf.predict(X_test)

accuracy_score(y_test, pred)
0.7908163265306123
data = titanic.iloc[:, [1, 6, 7, 11, 12, 13, 14, 15]]
# data
scaler = RobustScaler()
scaler.fit(data)
data2 = scaler.transform(data)
target = titanic.survived
pd.DataFrame(data2)
  0 1 2 3 4 5 6 7
0 -2.0 0.0 0.0 -1.0 0.0 -1.000000 0.0 -2.0
1 -2.0 1.0 2.0 0.0 0.0 -1.333333 3.0 -1.0
2 -2.0 1.0 2.0 -1.0 0.0 -1.000000 3.0 -1.0
3 -2.0 1.0 2.0 0.0 0.0 0.000000 0.0 -1.0
4 -2.0 1.0 2.0 -1.0 0.0 0.333333 0.0 -1.0
... ... ... ... ... ... ... ... ...
1301 0.0 1.0 0.0 -1.0 -2.0 -1.000000 1.0 0.0
1302 0.0 1.0 0.0 -1.0 -2.0 -1.000000 0.0 0.0
1303 0.0 0.0 0.0 0.0 -2.0 0.000000 0.0 0.0
1304 0.0 0.0 0.0 0.0 -2.0 0.000000 0.0 0.0
1305 0.0 0.0 0.0 0.0 0.0 0.000000 0.0 0.0

1306 rows × 8 columns

X_train, X_test, y_train, y_test = train_test_split(data2, target, train_size = 0.7,
                stratify=target)

rfclf = RandomForestClassifier()
rfclf.fit(X_train, y_train)
pred = rfclf.predict(X_test)

accuracy_score(y_test, pred)
0.7780612244897959
rfclf = DecisionTreeClassifier()
scores = cross_val_score(rfclf, data2, target, scoring='accuracy', cv = 10)
rfclf = RandomForestClassifier()
scores = cross_val_score(rfclf, data2, target, scoring='accuracy', cv = 10)
np.mean(scores)
0.7433529066353495
728x90
저작자표시

'PYTHON > 데이터분석' 카테고리의 다른 글

[머신러닝] 07. ROC 그래프  (0) 2023.01.03
[머신러닝] 06. 오차행렬  (0) 2023.01.03
[머신러닝] 04. 머신러닝 타이타닉 생존자 예측  (0) 2023.01.03
[머신러닝] 03. 훈련과 테스트 데이터  (0) 2023.01.03
[머신러닝] 02. sklearn  (0) 2023.01.02
'PYTHON/데이터분석' Related Articles more
Comments