240202 FRI 머신러닝 기초 실습

 

 

✅ 선형회귀 실습: 키 - 체중

 

In [ ]:

import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt 
import seaborn as sns 
import sklearn 

선형회귀 실습¶

1. 임의 데이터 생성

In [ ]:

heights = [187, 174, 179, 192, 188, 160, 179, 168, 168, 174]
weights = [87, 81, 82, 92, 90, 61, 86, 66, 69, 69]

body_df = pd.DataFrame({'height':heights, 'weight':weights})
body_df.head(3)

Out[ ]:

	height	weight
0	187	87
1	174	81
2	179	82

2. 데이터 시각화: 산점도

In [ ]:

sns.scatterplot(data=body_df, x='weight', y='height')
plt.title('Weight vs Height')
plt.xlabel('weight(kg)')
plt.ylabel('height(cm)')
plt.show()

3. 선형회귀 훈련(적합)

In [ ]:

from sklearn.linear_model import LinearRegression

#선형 모델을 가져와 변수에 저장
model_lr = LinearRegression()

# 타입 확인해보기
type(model_lr)

Out[ ]:

sklearn.linear_model._base.LinearRegression

In [ ]:

# DataFrame[] : Series(데이터 프레임의 컬럼)
# DataFrame[[]] : DataFrame 그대로

x = body_df[['weight']]
y = body_df[['height']]

In [ ]:

# 데이터 훈련
model_lr.fit(X = x, y = y)

Out[ ]:

LinearRegression()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

In [ ]:

# 가중치(w1) 확인, 변수에 지정
print(model_lr.coef_)
w1 = model_lr.coef_[0][0]

# 편향(bias, w0) 확인
print(model_lr.intercept_)
w0 = model_lr.intercept_[0]

[[0.86251245]]
[109.36527488]

In [ ]:

print('y = {}x + {}'.format(w1.round(2), w0.round(2)))

y = 0.86x + 109.37

결론: 키(height, y)는 몸무게(weight, x)에 0.86을 곱한 뒤 109.37을 더하면 된다.

4. 위 함수식을 활용하여 예측 컬럼 추가

In [ ]:

body_df['pred'] = body_df['weight']*w1 + w0
body_df.head(3)

Out[ ]:

	height	weight	pred
0	187	87	184.403858
1	174	81	179.228784
2	179	82	180.091296

5. 에러 계산 후 제곱

In [ ]:

body_df['error'] = body_df['height'] - body_df['pred']
body_df['error^2'] = body_df['error']*body_df['error']

body_df.head(3)

Out[ ]:

	height	weight	pred	error	error^2
0	187	87	184.403858	2.596142	6.739951
1	174	81	179.228784	-5.228784	27.340178
2	179	82	180.091296	-1.091296	1.190927

6. MSE 계산 (전부 더하기)

In [ ]:

body_df['error^2'].sum() / len(body_df)

Out[ ]:

10.152939045376309

7. 시각화

In [ ]:

sns.scatterplot(data = body_df, x ='weight', y = 'height')
sns.lineplot(data = body_df, x = 'weight', y = 'pred', color = 'red')

Out[ ]:

<Axes: xlabel='weight', ylabel='height'>

8. 모델 평가

In [ ]:

from sklearn.metrics import mean_squared_error, r2_score

In [ ]:

# 평가함수는 공통적으로 정답(실제 true), 예측값(pred)

y_true = body_df['height']
y_pred = body_df['pred']

mean_squared_error(y_true, y_pred)

Out[ ]:

10.152939045376309

In [ ]:

r2_score(y_true, y_pred)

Out[ ]:

0.8899887415172141

In [ ]:

y_pred2 = model_lr.predict(body_df[['weight']])
y_pred2

Out[ ]:

array([[184.40385835],
       [179.22878362],
       [180.09129608],
       [188.71642061],
       [186.99139571],
       [161.97853455],
       [183.54134589],
       [166.29109682],
       [168.87863418],
       [168.87863418]])

In [ ]:

mean_squared_error(y_true, y_pred2)

Out[ ]:

10.152939045376309

 

 

 

 

✅ 선형회귀 실습: 총 결제금액 - 팁

 

In [ ]:

import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt 
import seaborn as sns 
import sklearn 

C:\Users\LOVE\AppData\Local\Temp\ipykernel_17864\1283938912.py:2: DeprecationWarning: 
Pyarrow will become a required dependency of pandas in the next major release of pandas (pandas 3.0),
(to allow more performant data types, such as the Arrow string type, and better interoperability with other libraries)
but was not found to be installed on your system.
If this would cause problems for you,
please provide us feedback at https://github.com/pandas-dev/pandas/issues/54466
        
  import pandas as pd

선형회귀 실습 2¶

1. seaborn 내장 데이터셋 불러오기

In [ ]:

tips_df = sns.load_dataset('tips')
tips_df.head(3)

Out[ ]:

	total_bill	tip	sex	smoker	day	time	size
0	16.99	1.01	Female	No	Sun	Dinner	2
1	10.34	1.66	Male	No	Sun	Dinner	3
2	21.01	3.50	Male	No	Sun	Dinner	3

2. 데이터 시각화: 산점도

In [ ]:

sns.scatterplot(data = tips_df, x = 'total_bill', y = 'tip')
plt.title('total bill & tip')
plt.xlabel('total bill')
plt.ylabel('tip')
plt.show()

3. 선형회귀 훈련(적합)

In [ ]:

from sklearn.linear_model import LinearRegression
model_lr = LinearRegression()

x = tips_df[['total_bill']]
y = tips_df[['tip']]

model_lr.fit(x, y)

Out[ ]:

LinearRegression()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

In [ ]:

# 가중치(w1) 확인, 변수에 지정
print(model_lr.coef_)
w1 = model_lr.coef_[0][0]

# 편향(bias, w0) 확인
print(model_lr.intercept_)
w0 = model_lr.intercept_[0]

# 수식 확인
print('y = {}x + {}'.format(w1.round(2), w0.round(2)))

[[0.10502452]]
[0.92026961]
y = 0.11x + 0.92

결론: 총결제금액이 1$ 오를 때, 팁은 0.11$ 추가된다.

4. 위 함수식을 활용하여 예측값 추가

In [ ]:

y_true = tips_df['tip']
y_pred = model_lr.predict(tips_df[['total_bill']])

5. 모델 평가

In [ ]:

from sklearn.metrics import mean_squared_error, r2_score

mean_squared_error(y_true, y_pred)

Out[ ]:

1.036019442011377

In [ ]:

r2_score(y_true, y_pred)

Out[ ]:

0.45661658635167657

6. 시각화

In [ ]:

tips_df['pred'] = y_pred
tips_df.head(3)

Out[ ]:

	total_bill	tip	sex	smoker	day	time	size	pred
0	16.99	1.01	Female	No	Sun	Dinner	2	2.704636
1	10.34	1.66	Male	No	Sun	Dinner	3	2.006223
2	21.01	3.50	Male	No	Sun	Dinner	3	3.126835

In [ ]:

sns.scatterplot(data = tips_df, x ='total_bill', y = 'tip')
sns.lineplot(data = tips_df, x = 'total_bill', y = 'pred', color = 'red')

Out[ ]:

<Axes: xlabel='total_bill', ylabel='tip'>

성별, 흡연 유무 등 다른 변수에 대한 추가 분석 필요 → 다중선형회귀 분석 진행

7. 범주형 데이터 사용하기

In [ ]:

def get_sex(x):
    if x == 'Female':
        return 0
    else:
        return 1
    
tips_df['sex_en'] = tips_df['sex'].apply(get_sex)
tips_df.head(3)

Out[ ]:

	total_bill	tip	sex	smoker	day	time	size	pred	sex_en
0	16.99	1.01	Female	No	Sun	Dinner	2	2.704636	0
1	10.34	1.66	Male	No	Sun	Dinner	3	2.006223	1
2	21.01	3.50	Male	No	Sun	Dinner	3	3.126835	1

8. 훈련

In [ ]:

model_lr2 = LinearRegression()

x = tips_df[['total_bill', 'sex_en']]
y = tips_df[['tip']]

model_lr2.fit(x, y)

Out[ ]:

LinearRegression()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

9. 예측

In [ ]:

y_pred2 = model_lr2.predict(x)

10. 평가

In [ ]:

# 단순선형회귀 mse (x = total bill)
print('단순선형회귀', mean_squared_error(y_true, y_pred))

# 다중선형회귀 mse (x = total bill, sex)
print('다중선형회귀', mean_squared_error(y_true, y_pred2))

단순선형회귀 1.036019442011377
다중선형회귀 1.0358604137213616

In [ ]:

# 단순선형회귀 mse (x = total bill)
print('단순선형회귀', r2_score(y_true, y_pred))

# 다중선형회귀 mse (x = total bill, sex)
print('다중선형회귀', r2_score(y_true, y_pred2))

단순선형회귀 0.45661658635167657
다중선형회귀 0.45669999534149963

 

 

 

 

✅ 로지스틱회귀 실습: 타이타닉

 

In [ ]:

import numpy as np 
import pandas as pd 
import matplotlib.pyplot as plt 
import seaborn as sns 
import sklearn 

로지스틱 회귀 실습¶

1. 데이터셋 불러오기

In [ ]:

titanic_df = pd.read_csv('C:/Users/LOVE/Downloads/vscode/ML/titanic/train.csv', encoding='UTF-8')

titanic_df.head(3)

Out[ ]:

	PassengerId	Survived	Pclass	Name	Sex	Age	SibSp	Parch	Ticket	Fare	Cabin	Embarked
0	1	0	3	Braund, Mr. Owen Harris	male	22.0	1	0	A/5 21171	7.2500	NaN	S
1	2	1	1	Cumings, Mrs. John Bradley (Florence Briggs Th...	female	38.0	1	0	PC 17599	71.2833	C85	C
2	3	1	3	Heikkinen, Miss. Laina	female	26.0	0	0	STON/O2. 3101282	7.9250	NaN	S

2. 결측치 확인

In [ ]:

titanic_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 891 entries, 0 to 890
Data columns (total 12 columns):
 #   Column       Non-Null Count  Dtype  
---  ------       --------------  -----  
 0   PassengerId  891 non-null    int64  
 1   Survived     891 non-null    int64  
 2   Pclass       891 non-null    int64  
 3   Name         891 non-null    object 
 4   Sex          891 non-null    object 
 5   Age          714 non-null    float64
 6   SibSp        891 non-null    int64  
 7   Parch        891 non-null    int64  
 8   Ticket       891 non-null    object 
 9   Fare         891 non-null    float64
 10  Cabin        204 non-null    object 
 11  Embarked     889 non-null    object 
dtypes: float64(2), int64(5), object(5)
memory usage: 83.7+ KB

3. x = Fare

In [ ]:

x1 = titanic_df[['Fare']]
y_true = titanic_df[['Survived']]

from sklearn.linear_model import LogisticRegression
model_lor = LogisticRegression()
model_lor.fit(x1, y_true)

c:\Users\LOVE\AppData\Local\Programs\Python\Python39\lib\site-packages\sklearn\utils\validation.py:1229: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().
  y = column_or_1d(y, warn=True)

Out[ ]:

LogisticRegression()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

In [ ]:

sns.scatterplot(titanic_df, x = 'Fare', y = 'Survived')

Out[ ]:

<Axes: xlabel='Fare', ylabel='Survived'>

In [ ]:

sns.histplot(titanic_df, x = 'Fare')

Out[ ]:

<Axes: xlabel='Fare', ylabel='Count'>

In [ ]:

# 기술통계 (수치형)
titanic_df.describe()

Out[ ]:

	PassengerId	Survived	Pclass	Age	SibSp	Parch	Fare
count	891.000000	891.000000	891.000000	714.000000	891.000000	891.000000	891.000000
mean	446.000000	0.383838	2.308642	29.699118	0.523008	0.381594	32.204208
std	257.353842	0.486592	0.836071	14.526497	1.102743	0.806057	49.693429
min	1.000000	0.000000	1.000000	0.420000	0.000000	0.000000	0.000000
25%	223.500000	0.000000	2.000000	20.125000	0.000000	0.000000	7.910400
50%	446.000000	0.000000	3.000000	28.000000	0.000000	0.000000	14.454200
75%	668.500000	1.000000	3.000000	38.000000	1.000000	0.000000	31.000000
max	891.000000	1.000000	3.000000	80.000000	8.000000	6.000000	512.329200

In [ ]:

def get_att(x):
    print('클래스 종류', x.classes_)
    print('독립변수 개수', x.n_features_in_)
    print('들어간 독립변수(x)의 이름', x.feature_names_in_)
    print('가중치', x.coef_)
    print('바이어스', x.intercept_)

get_att(model_lor)

클래스 종류 [0 1]
독립변수 개수 1
들어간 독립변수(x)의 이름 ['Fare']
가중치 [[0.01519617]]
바이어스 [-0.94129222]

In [ ]:

from sklearn.metrics import accuracy_score, f1_score 

def get_metrics(true, pred):
    print('정확도', accuracy_score(true, pred))
    print('f1-score', f1_score(true, pred))
    

In [ ]:

y_pred_1 = model_lor.predict(x1)

y_pred_1[:10]

Out[ ]:

array([0, 1, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int64)

In [ ]:

get_metrics(y_true, y_pred_1)

정확도 0.6655443322109988
f1-score 0.354978354978355

결론: Fare가 유의미한 변수는 아님

4. 다중로지스틱회귀

In [ ]:

def get_sex(x):
    if x == 'female':
        return 0
    else:
        return 1

titanic_df['Sex_en'] = titanic_df['Sex'].apply(get_sex)
titanic_df.head(3)

Out[ ]:

	PassengerId	Survived	Pclass	Name	Sex	Age	SibSp	Parch	Ticket	Fare	Cabin	Embarked	Sex_en
0	1	0	3	Braund, Mr. Owen Harris	male	22.0	1	0	A/5 21171	7.2500	NaN	S	1
1	2	1	1	Cumings, Mrs. John Bradley (Florence Briggs Th...	female	38.0	1	0	PC 17599	71.2833	C85	C	0
2	3	1	3	Heikkinen, Miss. Laina	female	26.0	0	0	STON/O2. 3101282	7.9250	NaN	S	0

In [ ]:

x2 = titanic_df[['Pclass', 'Sex_en', 'Fare']]
y_true = titanic_df[['Survived']]

model_lor2 = LogisticRegression()
model_lor2.fit(x2, y_true)

c:\Users\LOVE\AppData\Local\Programs\Python\Python39\lib\site-packages\sklearn\utils\validation.py:1229: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().
  y = column_or_1d(y, warn=True)

Out[ ]:

LogisticRegression()

In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.

In [ ]:

get_att(model_lor2)

클래스 종류 [0 1]
독립변수 개수 3
들어간 독립변수(x)의 이름 ['Pclass' 'Sex_en' 'Fare']
가중치 [[-8.88331324e-01 -2.53993425e+00  1.64019087e-03]]
바이어스 [3.02004403]

In [ ]:

y_pred_2 = model_lor2.predict(x2)

y_pred_2[:10]

Out[ ]:

array([0, 1, 1, 1, 0, 0, 0, 0, 1, 1], dtype=int64)

In [ ]:

print('# X: Fare')
get_metrics(y_true, y_pred_1)

print('# X: Fare, Pclass, Sex')
get_metrics(y_true, y_pred_2)

# X: Fare
정확도 0.6655443322109988
f1-score 0.354978354978355
# X: Fare, Pclass, Sex
정확도 0.7867564534231201
f1-score 0.7121212121212122

In [ ]:

# 각 데이터별 Y=1일 확률 (= 생존할 확률)
model_lor2.predict_proba(x2)

Out[ ]:

array([[0.8977979 , 0.1022021 ],
       [0.09546762, 0.90453238],
       [0.40901264, 0.59098736],
       ...,
       [0.40287202, 0.59712798],
       [0.58880217, 0.41119783],
       [0.89772263, 0.10227737]])