[ML] Kaggle_뇌졸중 데이터를 활용한 분류모델 만들기(불균형데이터,SMOTE)

 

Cerebral Stroke Prediction-Imbalanced Dataset

1. 라이브러리 불러오기

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.patches as mpatches

from sklearn.impute import KNNImputer #Imputation
%matplotlib inline

import warnings
warnings.filterwarnings("ignore")

---

2. 데이터 확인

df = pd.read_csv('dataset.csv')
df.info()

import seaborn as sns
import matplotlib.pyplot as plt

sns.countplot(x='stroke',data=df)
plt.title("Imbalanced data")
plt.show()

print('Target class is ', '{0:0.4f}'. format(783/(42617+783)*100), '%')

<Target class is 1.8041 %>

---

3. 데이터 전처리

• 인코딩 - 원핫
• 결측값 처리 - KNNInputer
• 스케일링 - 정규화

# 인코딩
df = pd.get_dummies(df,columns=['gender','ever_married','work_type','Residence_type','smoking_status'])

# 결측값 처리 
imputer = KNNImputer(missing_values=np.nan)
tab = imputer.fit_transform(df)
df_new = pd.DataFrame(tab, columns=df.columns)

# target 분리
X = df_new.drop('stroke',axis=1)
y = df_new['stroke']

# 스케일링
from sklearn.preprocessing import MinMaxScaler
MM = MinMaxScaler()
X = MM.fit_transform(X)

---

4. 데이터셋 분리

from sklearn.model_selection import train_test_split
X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.2,random_state=1)

print("X_train.shape : ", X_train.shape)
print("y_train.shape : ", y_train.shape)
print("X_test.shape : " ,X_test.shape)
print("y_test.shape : " ,y_test.shape)

---

5. 모델링

5-1. 불균형 데이터 그대로 

from sklearn.metrics import classification_report
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import GradientBoostingClassifier
from lightgbm import LGBMClassifier

dt = DecisionTreeClassifier()
rf = RandomForestClassifier()
lr = LogisticRegression()
kc = KNeighborsClassifier()
svc = SVC()
gbc = GradientBoostingClassifier()
lgv = LGBMClassifier()

models = [dt,rf,lr,kc,svc,gbc,lgv]

for model in models:
    print("MODEL NAME:",model)
    model.fit(X_train,y_train)
    y_pred = model.predict(X_test)
    
    print(classification_report(y_test,y_pred))

5-2. 오버샘플링

from imblearn.over_sampling import SMOTE

sm = SMOTE(random_state=1)
X_sm,y_sm = sm.fit_resample(X,y)

X_train,X_test,y_train,y_test = train_test_split(X_sm,y_sm,test_size=0.2,random_state=1)

dt1 = DecisionTreeClassifier()
rf1 = RandomForestClassifier()
lr1 = LogisticRegression()
kc1 = KNeighborsClassifier()
lgv1 = LGBMClassifier()

models = [dt1,rf1,lr1,kc1,lgv1]

for model in models:
    print("MODEL NAME:",model)
    model.fit(X_train,y_train)
    y_pred = model.predict(X_test)
    
    print(classification_report(y_test,y_pred))

5-3. 언더샘플링

from imblearn.under_sampling import RandomUnderSampler

us = RandomUnderSampler(random_state=1)
X_us,y_us = us.fit_resample(X,y)

X_train,X_test,y_train,y_test = train_test_split(X_us,y_us,test_size=0.2,random_state=1)

dt2 = DecisionTreeClassifier()
rf2 = RandomForestClassifier()
lr2 = LogisticRegression()
kc2 = KNeighborsClassifier()
lgv2 = LGBMClassifier()

models = [dt2,rf2,lr2,kc2,lgv2]

for model in models:
    print("MODEL NAME:",model)
    model.fit(X_train,y_train)
    y_pred = model.predict(X_test)
    
    print(classification_report(y_test,y_pred))

---

6. 최종 모델 = 오버샘플링한 모델