[zerobase_데이터취업스쿨] 머신러닝_CH7-04_Credit Fraud Detection(불균형 데이터분석)

CH7-04 Credit Card Fraud Detection 데이터셋 분석

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import koreanize_matplotlib
from sklearn.metrics import accuracy_score, f1_score, recall_score, precision_score, roc_curve, roc_auc_score
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier
from xgboost import XGBClassifier
from lightgbm import LGBMClassifier

raw_df = pd.read_csv('../datas/creditcard.csv')
raw_df

raw_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 284807 entries, 0 to 284806
Data columns (total 31 columns):
 #   Column  Non-Null Count   Dtype  
---  ------  --------------   -----  
 0   Time    284807 non-null  float64
 1   V1      284807 non-null  float64
 2   V2      284807 non-null  float64
 3   V3      284807 non-null  float64
 4   V4      284807 non-null  float64
 5   V5      284807 non-null  float64
 6   V6      284807 non-null  float64
 7   V7      284807 non-null  float64
 8   V8      284807 non-null  float64
 9   V9      284807 non-null  float64
 10  V10     284807 non-null  float64
 11  V11     284807 non-null  float64
 12  V12     284807 non-null  float64
 13  V13     284807 non-null  float64
 14  V14     284807 non-null  float64
 15  V15     284807 non-null  float64
 16  V16     284807 non-null  float64
 17  V17     284807 non-null  float64
 18  V18     284807 non-null  float64
 19  V19     284807 non-null  float64
 20  V20     284807 non-null  float64
 21  V21     284807 non-null  float64
 22  V22     284807 non-null  float64
 23  V23     284807 non-null  float64
 24  V24     284807 non-null  float64
 25  V25     284807 non-null  float64
 26  V26     284807 non-null  float64
 27  V27     284807 non-null  float64
 28  V28     284807 non-null  float64
 29  Amount  284807 non-null  float64
 30  Class   284807 non-null  int64  
dtypes: float64(30), int64(1)
memory usage: 67.4 MB

1. 칼럼 명세

• V1~V28은 개인정보보호를 이유로 PCA로 차원축소된 개인정보 피처들임
• Amount는 거래량
• Class 1은 Fraud, 0은 Not Fraud

2. 4개 모델 한꺼번에 훈련하는 함수, [f1,precision,accuracy,recall] 스코어 함수, 그 점수를 데이터프레임으로 모델이름, 점수 담는 함수 만들기

def scores(pred, y_test):

    f1 = f1_score(y_test, pred)
    acc = accuracy_score(y_test, pred)
    precision = precision_score(y_test, pred)
    rec = recall_score(y_test, pred)
    roc_score = roc_auc_score(y_test, pred)
    
    return acc, precision, rec, f1, roc_score

def train_models(model, X_train, y_train, X_test, y_test):
    model.fit(X_train, y_train)     
    pred = model.predict(X_test)
    proba = model.predict_proba(X_test)
    
    return scores(pred, y_test), proba

def get_score_df(models, X_train, y_train, X_test, y_test):
    tmp = []
    names = ['LogisticRegression', 'RandomForest', 'XGBM', 'LGBM']
    for model,name in zip(models, names):
        
        score, proba = train_models(model, X_train, y_train, X_test, y_test)
        tmp.append(pd.DataFrame({'accuracy':score[0],
                                 'precision':score[1],
                                 'recall':score[2],
                                 'f1_score':score[3],
                                'roc_auc_score':score[4]},index = [name]))
        
        fpr, tpr, thresholds = roc_curve(y_test, proba[:,1])
        plt.plot([0,1],[0,1],color = 'red', ls = 'dashed')
        plt.plot(fpr, tpr, color = 'blue', label = f'{name}')
        plt.legend()
        plt.title('roc_auc_curve')
        plt.show()
        
    return pd.concat(tmp, axis = 0)

3. 클래스 불균형 살펴보기

sns.countplot(raw_df, x = 'Class', hue = 'Class')

<Axes: xlabel='Class', ylabel='count'>

raw_df.describe()

4. 클래스의 비율 value_counts(normalize=True)로 확인하기

• Fraud인 클래스 1의 비율이 전체에서 0.1727%로 굉장히 심하게 불균형한 데이터임

raw_df['Class'].value_counts(normalize=True)

Class
0    0.998273
1    0.001727
Name: proportion, dtype: float64

5. 일단 아무런 전처리 없이 모델 돌려보기

X = raw_df.drop(['Time','Class'], axis = 1)
y = raw_df['Class']

X_train, X_test, y_train, y_test = train_test_split(X, y, stratify = y, test_size = 0.2, random_state = 42)

lr_clf = LogisticRegression(n_jobs = -1, random_state = 42, solver = 'liblinear')
rf_clf = RandomForestClassifier(n_estimators = 300, oob_score = True, n_jobs = -1, random_state = 42)
XGB_clf = XGBClassifier(n_jobs = -1, random_state = 42, verbosity = 1, n_estimators = 100)
LGBM_clf = LGBMClassifier(n_jobs = -1, random_state = 42, verbosity = 1, n_estimators = 1000,num_leaves = 186, learning_rate=0.15, boost_from_average = False)

models = [lr_clf, rf_clf, XGB_clf, LGBM_clf]

res = get_score_df(models, X_train, y_train, X_test, y_test)

C:\Users\kd010\miniconda3\Lib\site-packages\sklearn\linear_model\_logistic.py:1222: UserWarning: 'n_jobs' > 1 does not have any effect when 'solver' is set to 'liblinear'. Got 'n_jobs' = 16.
  warnings.warn(

[LightGBM] [Info] Number of positive: 394, number of negative: 227451
[LightGBM] [Info] Auto-choosing col-wise multi-threading, the overhead of testing was 0.009062 seconds.
You can set `force_col_wise=true` to remove the overhead.
[LightGBM] [Info] Total Bins 7395
[LightGBM] [Info] Number of data points in the train set: 227845, number of used features: 29
[LightGBM] [Warning] No further splits with positive gain, best gain: -inf

res