paddle 练习（二）使用基础API完成模型训练&预测

模型训练：

下面展示模型训练的代码。

这里用到的是线性回归模型最常用的损失函数–均方误差（MSE），用来衡量模型预测的房价和真实房价的差异。

对损失函数进行优化所采用的方法是梯度下降法.

# 将训练数据集和测试数据集按照8:2的比例分开
ratio = 0.8
offset = int(housing_data.shape[0] * ratio)
train_data = housing_data[:offset]
test_data = housing_data[offset:]


import paddle.nn.functional as F 
y_preds = []
labels_list = []

def train(model):
    print('start training ... ')
    # 开启模型训练模式
    model.train()
    EPOCH_NUM = 500
    train_num = 0
    optimizer = paddle.optimizer.SGD(learning_rate=0.001, parameters=model.parameters())
    for epoch_id in range(EPOCH_NUM):
        # 在每轮迭代开始之前，将训练数据的顺序随机的打乱
        np.random.shuffle(train_data)
        # 将训练数据进行拆分，每个batch包含20条数据
        mini_batches = [train_data[k: k+BATCH_SIZE] for k in range(0, len(train_data), BATCH_SIZE)]
        for batch_id, data in enumerate(mini_batches):
            features_np = np.array(data[:, :13], np.float32)
            labels_np = np.array(data[:, -1:], np.float32)
            features = paddle.to_tensor(features_np)
            labels = paddle.to_tensor(labels_np)
            # 前向计算
            y_pred = model(features)
            cost = F.mse_loss(y_pred, label=labels)
            train_cost = cost.numpy()[0]
            # 反向传播
            cost.backward()
            # 最小化loss，更新参数
            optimizer.step()
            # 清除梯度
            optimizer.clear_grad()
            
            if batch_id%30 == 0 and epoch_id%50 == 0:
                print("Pass:%d,Cost:%0.5f"%(epoch_id, train_cost))

            train_num = train_num + BATCH_SIZE
            train_nums.append(train_num)
            train_costs.append(train_cost)
        
model = Regressor()
train(model)



matplotlib.use('TkAgg')
# matplotlib inline
draw_train_process(train_nums, train_costs)

如果你想成功运行这段代码，请参考我的paddle练习（一）种的开始数据集house.data加载部分代码。

paddle练习（一）使用线性回归预测波士顿房价_Vertira的博客-CSDN博客import paddleimport numpy as npimport osimport matplotlibimport matplotlib.pyplot as pltimport pandas as pdimport seaborn as snsimport warningswarnings.filterwarnings("ignore")print(paddle.__version__)# 从文件导入数据datafile = 'housing.data'housin.https://blog.csdn.net/Vertira/article/details/122171950

运行结果：

start training ... 
Pass:0,Cost:507.42090
Pass:50,Cost:47.54215
Pass:100,Cost:83.45570
Pass:150,Cost:86.61785
Pass:200,Cost:32.05870
Pass:250,Cost:15.67683
Pass:300,Cost:23.19898
Pass:350,Cost:48.89576
Pass:400,Cost:56.87611
Pass:450,Cost:33.11672

然后进行模型预测

# 获取预测数据
INFER_BATCH_SIZE = 100

infer_features_np = np.array([data[:13] for data in test_data]).astype("float32")
infer_labels_np = np.array([data[-1] for data in test_data]).astype("float32")

infer_features = paddle.to_tensor(infer_features_np)
infer_labels = paddle.to_tensor(infer_labels_np)
fetch_list = model(infer_features)

sum_cost = 0
for i in range(INFER_BATCH_SIZE):
    infer_result = fetch_list[i][0]
    ground_truth = infer_labels[i]
    if i % 10 == 0:
        print("No.%d: infer result is %.2f,ground truth is %.2f" % (i, infer_result, ground_truth))
    cost = paddle.pow(infer_result - ground_truth, 2)
    sum_cost += cost
mean_loss = sum_cost / INFER_BATCH_SIZE
print("Mean loss is:", mean_loss.numpy())

预测结果：

No.0: infer result is 11.91,ground truth is 8.50
No.10: infer result is 5.13,ground truth is 7.00
No.20: infer result is 14.46,ground truth is 11.70
No.30: infer result is 16.31,ground truth is 11.70
No.40: infer result is 13.45,ground truth is 10.80
No.50: infer result is 15.81,ground truth is 14.90
No.60: infer result is 18.66,ground truth is 21.40
No.70: infer result is 15.23,ground truth is 13.80
No.80: infer result is 17.96,ground truth is 20.60
No.90: infer result is 21.41,ground truth is 24.50

画图显示：

def plot_pred_ground(pred, ground):
    plt.figure()   
    plt.title("Predication v.s. Ground truth", fontsize=24)
    plt.xlabel("ground truth price(unit:$1000)", fontsize=14)
    plt.ylabel("predict price", fontsize=14)
    plt.scatter(ground, pred, alpha=0.5)  #  scatter:散点图,alpha:"透明度"
    plt.plot(ground, ground, c='red')
    plt.show()

欢迎点赞，收藏，关注。

plot_pred_ground(fetch_list, infer_labels_np)

欢迎点赞，收藏，关注。