스터디노트 (Cost Function 2)

📌 Boston 집 값에 대해 자세하게 알아보자

boston.head()

#집 값에 대한 히스토그램
import plotly_express as px

fig = px.histogram(boston, x='PRICE')
fig.show()

📌 상관 계수 구하기

import matplotlib.pyplot as plt
import seaborn as sns

corr_mat = boston.corr().round(1)
sns.heatmap(data = corr_mat, annot=True, cmap='bwr')

📌 plot 그림 그리기 

sns.set_style('darkgrid')
sns.set(rc={'figure.figsize':(12, 6)})

fig, ax = plt.subplots(ncols = 2)
sns.regplot(x="RM", y="PRICE", data = boston, ax=ax[0])
sns.regplot(x="LSTAT", y="PRICE", data = boston, ax=ax[1])

📌sklearn

from sklearn.model_selection import train_test_split

X = boston.drop('PRICE', axis=1)
y = boston['PRICE']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=13)

from sklearn.linear_model import LinearRegression

reg = LinearRegression()
reg.fit(X_train, y_train)

from sklearn.metrics import mean_squared_error
import numpy as np
pred_tr = reg.predict(X_train)
pred_test = reg.predict(X_test)
rmse_tr = (np.sqrt(mean_squared_error(y_train, pred_tr)))
rmse_test = (np.sqrt(mean_squared_error(y_test, pred_test)))

print('RMSE Train data : ', rmse_tr)
print('RMSE Test data : ', rmse_test)

plt.scatter(y_test, pred_test) #참 값과 테스트 데이터에 대한 값
plt.xlabel('Real ($1000)')
plt.ylabel('Predicted Prices')
plt.plot([0, 50], [0, 50])
plt.show()

---

📌sklearn 

- LSTAT 값도 같이 빼고 모델을 만들어보자.

X = boston.drop(['PRICE', 'LSTAT'], axis=1)
y = boston['PRICE']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=13)

reg = LinearRegression()
reg.fit(X_train, y_train)

pred_tr = reg.predict(X_train)
pred_test = reg.predict(X_test)
rmse_tr = (np.sqrt(mean_squared_error(y_train, pred_tr)))
rmse_test = (np.sqrt(mean_squared_error(y_test, pred_test)))

print('RMSE Train data : ', rmse_tr)
print('RMSE Test data : ', rmse_test)
>>>>
RMSE Train data :  5.165137874244864
RMSE Test data :  5.2955950325971655

plt.scatter(y_test, pred_test)
plt.xlabel('Real ($1000)')
plt.ylabel('Predicted Prices')
plt.plot([0, 50], [0, 50])
plt.show()