梯度下降的优化方法对波士顿房价进行预测
def linear2(): """ 梯度下降的优化方法对波士顿房价进行预测 :return: """ # 1)获取数据 boston = load_boston() print("特征数量:\n", boston.data.shape) # 2)划分数据集 x_train, x_test, y_train, y_test = train_test_split(boston.data, boston.target, random_state=22) # 3)标准化 transfer = StandardScaler() x_train = transfer.fit_transform(x_train) x_test = transfer.transform(x_test) # 4)预估器 estimator = SGDRegressor(learning_rate="constant", eta0=0.01, max_iter=10000, penalty="l1") estimator.fit(x_train, y_train) # 5)得出模型 print("梯度下降-权重系数为:\n", estimator.coef_) print("梯度下降-偏置为:\n", estimator.intercept_) # 6)模型评估 y_predict = estimator.predict(x_test) print("预测房价:\n", y_predict) error = mean_squared_error(y_test, y_predict) print("梯度下降-均方误差为:\n", error) return None