California Housing




1. 1. Linear Regression


Prédictions
X, y = fetch_california_housing(return_X_y=True)

model = LinearRegression()
model.fit(X, y)

predictions = model.predict(X)
[4.13164983 3.97660644 3.67657094 ... 0.17125141 0.31910524 0.51580363]

Valeurs attendues (en 100 000 $)
y
[4.526 3.585 3.521 ... 0.923 0.847 0.894]



2. 2. Différences


Différences
differences = predictions - y
[-0.39435017  0.39160644  0.15557094 ... -0.75174859 -0.52789476
 -0.37819637]
Histogramme des différences
import matplotlib.pyplot as plt
plt.hist(differences, bins=30)


MAE (Mean Absolute Error) = Moyenne des valeurs absolues des différences
result = np.mean(np.abs(differences))
0.5311643817546456
Boxplot des différences
fig = plt.figure(figsize =(10, 4))
plt.boxplot(differences, vert=False)




3. 3. MAE (Mean Absolute Error)


MAE
from sklearn.metrics import mean_absolute_error
result = mean_absolute_error(y_true=y, y_pred=predictions)
0.5311643817546456



4. 4. MSE (Mean Squared Error)


MSE
from sklearn.metrics import mean_squared_error
result = mean_squared_error(y_true=y, y_pred=predictions)
0.5243209861846072



5. 5. RMSE (Root Mean Square Error)


RMSE
from sklearn.metrics import root_mean_squared_error
result = root_mean_squared_error(y_true=y, y_pred=predictions)
0.7241001216576387