Um simples algoritmo para implementar o Ridge Regression deve ser usado importando o Ridge
do sklearn.linear_model
, e assim, basta determinar os parâmetros do modelo.
import matplotlib.pyplot as plt
from sklearn.linear_model import Ridge
def ridge_regression(data, predictors, alpha, models_to_plot={}):
RidgeReg = Ridge(alpha=alpha,normalize=True)
RidgeReg.fit(dados[X],dados['y'])
y_pred = ridgereg.predict(dados[X])
if alpha in ridplot:
plt.subplot(ridplot[alpha])
plt.tight_layout()
plt.plot(dados['x'],y_pred)
plt.plot(dados['x'],data['y'],'.')
plt.title('Plot for alpha: %.3g'%alpha)
resid2 = sum((y_pred-data['y'])**2)
ret = [resid2]
ret.extend([RidgeReg.intercept_])
ret.extend(RidgeReg.coef_)
return ret