[머신러닝 심화 2주차]

 

 

 

import pandas as pd

import matplotlib.pyplot as plt

from sklearn.preprocessing import LabelEncoder

from sklearn.tree import DecisionTreeClassifier,plot_tree

titaninc_df = pd.read_csv('C:/Users/82106/Desktop/데이터분석/강의/6.파이썬/4.머신러닝/ML/titanic/train.csv')

titaninc_df.info()

 

 

 

#Pclass: LabelEncoder

#Sex: LabelEncoder

#Age: 결측치-> 평균으로 대치하고

le = LabelEncoder()

titaninc_df['Sex'] = le.fit_transform(titaninc_df['Sex'])

le2 = LabelEncoder()

titaninc_df['Pclass'] = le2.fit_transform(titaninc_df['Pclass'])

age_mean = titaninc_df['Age'].mean()

titaninc_df['Age'] = titaninc_df['Age'].fillna(age_mean)

le3 = LabelEncoder()

titaninc_df['Embarked'] = titaninc_df['Embarked'].fillna('S')

titaninc_df['Embarked'] = le3.fit_transform(titaninc_df['Embarked'])

X = titaninc_df[X_features]

y = titaninc_df['Survived']

model_dt = DecisionTreeClassifier(max_depth = 1)

model_dt.fit(X,y)

plt.figure(figsize = (10,5))

plot_tree(model_dt, feature_names=X_features, class_names=['Not Survived','Survived'], filled= True)

plt.show()

 

 

 

저 DecisionTreeClassifier  옆에 (max_depth =  ) 숫자가  몇 개 보일지 정하는 것!!

 

max_depth =  1 일 때 

 

( ) 아예 저 괄호 안을 비워놨을 때  

 

 

 

# 로지스틱회귀, 의사결정나무, 랜덤포레스트

from sklearn.linear_model import LogisticRegression

from sklearn.tree import DecisionTreeClassifier

from sklearn.ensemble import RandomForestClassifier

from sklearn.metrics import accuracy_score,f1_score

model_lor = LogisticRegression()

model_dt = DecisionTreeClassifier(random_state=42)

model_rf = RandomForestClassifier(random_state=42)

model_lor.fit(X,y)

model_dt.fit(X,y)

model_rf.fit(X,y)

y_lor_pred = model_lor.predict(X)

y_dt_pred = model_dt.predict(X)

y_rf_pred = model_rf.predict(X)

def get_score(model_name, y_true, y_pred):

    acc = accuracy_score(y_true, y_pred).round(3)

    f1 = f1_score(y_true,y_pred).round(3)

    print(model_name, 'acc 스코어는: ',acc, 'f1_score는: ', f1)

   

get_score('lor',y,y_lor_pred)

get_score('dt ',y,y_dt_pred)

get_score('rf ',y,y_rf_pred)

 

 

X_features

 

 

model_rf.feature_importances_

 

 

# 최근접 이웃, 부스팅 모델 적용

 

# !pip install xgboost

# !pip install lightgbm

 

 

from sklearn.neighbors import KNeighborsClassifier

from sklearn.ensemble import GradientBoostingClassifier

from xgboost import XGBClassifier

from lightgbm import LGBMClassifier

model_knn = KNeighborsClassifier()

model_gbm = GradientBoostingClassifier(random_state= 42)

model_xgb = XGBClassifier(random_state= 42)

model_lgb = LGBMClassifier(random_state= 42)

model_knn.fit(X,y)

model_gbm.fit(X,y)

model_xgb.fit(X,y)

model_lgb.fit(X,y)

y_knn_pred = model_knn.predict(X)

y_gbm_pred = model_gbm.predict(X)

y_xgb_pred = model_xgb.predict(X)

y_lgb_pred = model_lgb.predict(X)

get_score('lor',y,y_lor_pred)

get_score('dt ',y,y_dt_pred)

get_score('rf ',y,y_rf_pred)

get_score('knn',y,y_knn_pred)

get_score('gbm ',y,y_gbm_pred)

get_score('xgb ',y,y_xgb_pred)

get_score('lgb ',y,y_lgb_pred)

 

 

acc (accuracy) 정확도 / f1- score : 정밀도