house-price-save.py

# 查看房价数据集
from keras.datasets import boston_housing

(train_data, train_targets), (test_data, test_targets) = boston_housing.load_data()

# 查看数据
print(train_data.shape)
print(test_data.shape)

print(train_targets)

# 正规化数据
mean = train_data.mean(axis=0)
train_data -= mean
std = train_data.std(axis=0)
train_data /= std

test_data -= mean
test_data /= std

# 构建网络，当训练数据很少时，很容易过拟合，使用一个简单的网络是避免过拟合的一个方式
from keras import models
from keras import layers


def build_model():
    # 我们需要多次使用相同的模型，所以使用一个函数去构建它
    model = models.Sequential()
    model.add(layers.Dense(64, activation='relu', input_shape=(train_data.shape[1],)))
    model.add(layers.Dense(64, activation='relu'))
    model.add(layers.Dense(1))
    # mse是Mean Squared Error
    # mae是Mean Absolute Error
    model.compile(optimizer='rmsprop',
                  loss='mse',
                  metrics=['mae'])
    return model


# k分区的交叉验证
import numpy as np

k = 4
num_val_samples = len(train_data) // k
num_epochs = 500
all_mae_histories = []
for i in range(k):
    print('processing fold #', i)
    # 准备分区i为验证数据
    val_data = train_data[i * num_val_samples: (i + 1) * num_val_samples]
    val_targets = train_targets[i * num_val_samples: (i + 1) * num_val_samples]

    # 准备其他的分区为训练数据
    partial_train_data = np.concatenate(
        [train_data[:i * num_val_samples],
         train_data[(i + 1) * num_val_samples:]],
        axis=0)
    partial_train_targets = np.concatenate(
        [train_targets[:i * num_val_samples],
         train_targets[(i + 1) * num_val_samples:]],
        axis=0)

    # 构建模型（已编译）
    model = build_model()
    # 训练模型
    history = model.fit(partial_train_data, partial_train_targets,
                        epochs=num_epochs, batch_size=1, verbose=0,
                        validation_data=(val_data, val_targets))
    mae_history = history.history['val_mean_absolute_error']
    all_mae_histories.append(mae_history)

# 计算每个轮回K分区交叉验证分数的平均值
average_mae_history = [np.mean([x[i] for x in all_mae_histories]) for i in range(num_epochs)]

# 画出验证分数
import matplotlib.pyplot as plt

plt.plot(range(1, len(average_mae_history) + 1), average_mae_history)
plt.xlabel('Epochs')
plt.ylabel('Validation MAE')
plt.show()


# 画平滑的曲线，去除前10个数据点
def smooth_curve(points, factor=0.9):
    smoothed_points = []
    for point in points:
        if smoothed_points:
            previous = smoothed_points[-1]
            smoothed_points.append(previous * factor + point * (1 - factor))
        else:
            smoothed_points.append(point)
    return smoothed_points


smooth_mae_history = smooth_curve(average_mae_history[10:])

plt.plot(range(1, len(smooth_mae_history) + 1), smooth_mae_history)
plt.xlabel('Epochs')
plt.ylabel('Validation MAE')
plt.show()

# 训练最终的模型
model = build_model()
# 使用全部的训练集
model.fit(train_data, train_targets,
          epochs=80, batch_size=16, verbose=0)
test_mse_score, test_mae_score = model.evaluate(test_data, test_targets)

print(test_mae_score)