线性回归算法
class LR_LS():
def __init__(self):
self.w = None
def fit(self, X, y):
# 最小二乘法矩阵求解
#============================= show me your code =======================
self.w = np.linalg.inv(X.T.dot(X)).dot(X.T).dot(y)
#============================= show me your code =======================
def predict(self, X):
# 用已经拟合的参数值预测新自变量
#============================= show me your code =======================
y_pred = X.dot(self.w)
#============================= show me your code =======================
return y_pred
if __name__ == "__main__":
lr_ls = LR_LS()
lr_ls.fit(x,y)
print("估计的参数值：%s" %(lr_ls.w))
x_test = np.array([2,4,5]).reshape(1,-1)
print("预测值为: %s" %(lr_ls.predict(x_test)))