當AI遇上股票
台南一中AI選修教材 / 顏永進

1: print(df[:5])

1: data = df.filter(['Close'])
2: print(data.head())

 1: import numpy as np
 2: 
 3: # 以前N天的股價來預測未來股價
 4: previousNDays = 10
 5: x_data, y_data = [], []
 6: for i in range(len(sc_data) - previousNDays):
 7:   x = sc_data[i:i+previousNDays]
 8:   y = sc_data[i+previousNDays]
 9:   x_data.append(x)
10:   y_data.append(y)
11: #因為等一下要送進tensorflow，所以先轉成numpy的陣列格式
12: x_data, y_data = np.array(x_data), np.array(y_data)
13: 
14: print(x_data.shape)
15: print(y_data.shape)

 1: ratio = 0.8
 2: train_size = round(len(x_data) * ratio)
 3: print(train_size)
 4: # 第0筆到第train_size-1筆的資料分割為訓練集
 5: x_train, y_train = x_data[:train_size], y_data[:train_size]
 6: # 第train_size筆到最後一筆的資料分割為測試集
 7: x_test, y_test = x_data[train_size:], y_data[train_size:]
 8: #from sklearn.model_selection import train_test_split
 9: #x_train, x_test, y_train, y_test = train_test_split(x_data, y_data, test_size=0.2)
10: print(x_train.shape)
11: print(y_train.shape)
12: print(x_test.shape)
13: print(y_test.shape)

 1: # CNN模型
 2: import tensorflow as tf
 3: from tensorflow.keras import layers
 4: 
 5: #建構CNN模型
 6: model = tf.keras.Sequential([
 7:     layers.Input(shape=(previousNDays, 1, 1)),
 8:     #卷積層：用 3 天的窗格掃描股價模式
 9:     layers.Conv2D(filters=512, kernel_size=(3, 1), activation='relu'),
10:     #攤平
11:     layers.Flatten(),
12:     #全連接層/輸出層（回歸問題不加激活函數）
13:     layers.Dense(1)
14: ])

1: model.summary()

1: model.compile(loss='mse', optimizer='adam', metrics=['mae'])

1: model.fit(x_train, y_train,
2:           validation_split=0.2,
3:           batch_size=200, epochs=20)

1: score = model.evaluate(x_test, y_test)
2: print('loss:', score[0])
3: print('mae:', score[1])

1: predict = model.predict(x_test)
2: predict = scaler.inverse_transform(predict)
3: predict = np.reshape(predict, (predict.size,))
4: ans = scaler.inverse_transform(y_test)
5: ans = np.reshape(ans, (ans.size,))
6: print(predict[:3])
7: print(ans[:3])

1: plt.clf()
2: plt.plot(predict, label='Predicted')
3: plt.plot(ans, label='Actual')
4: plt.legend()
5: plt.show()

1: print(df[:5])

1: data = df.filter(['Close', 'Volume'])
2: print(data.head())

 1: import numpy as np
 2: 
 3: # 以前N天的股價來預測未來股價
 4: previousNDays = 10
 5: x_data, y_data = [], []
 6: for i in range(len(scal_all_x) - previousNDays):
 7:   x = scal_all_x[i:i+previousNDays]
 8:   y = scal_all_y[i+previousNDays]
 9:   x_data.append(x)
10:   y_data.append(y)
11: #因為等一下要送進tensorflow
12: x_data, y_data = np.array(x_data), np.array(y_data)
13: 
14: print(x_data.shape)
15: print(y_data.shape)

 1: ratio = 0.8
 2: train_size = round(len(x_data) * ratio)
 3: print(train_size)
 4: x_train, y_train = x_data[:train_size], y_data[:train_size]
 5: x_test, y_test = x_data[train_size:], y_data[train_size:]
 6: 
 7: print(x_train.shape)
 8: print(y_train.shape)
 9: print(x_test.shape)
10: print(y_test.shape)

 1: # CNN模型
 2: import tensorflow as tf
 3: from tensorflow.keras import layers
 4: 
 5: #建構CNN模型
 6: model = tf.keras.Sequential([
 7:     layers.Input(shape=(previousNDays, 2, 1)),
 8:     #第一個卷積層
 9:     layers.Conv2D(filters=512, kernel_size=(3, 1), activation='relu'),
10:     #第二個卷積層
11:     layers.Conv2D(filters=512, kernel_size=(3, 1), activation='relu'),
12:     #攤平
13:     layers.Flatten(),
14:     #全連接層/輸出層（回歸問題不加激活函數）
15:     layers.Dense(1)
16: ])

1: model.summary()

1: model.compile(loss='mse', optimizer='adam', metrics=['mae'])

1: model.fit(x_train, y_train,
2:           validation_split=0.2,
3:           batch_size=200, epochs=20)

1: score = model.evaluate(x_test, y_test)
2: print('loss:', score[0])
3: print('mae:', score[1])

1: predict = model.predict(x_test)
2: predict = scalerY.inverse_transform(predict)
3: predict = np.reshape(predict, (predict.size,))
4: ans = scalerY.inverse_transform(y_test)
5: ans = np.reshape(ans, (ans.size,))
6: print(predict[:3])
7: print(ans[:3])

1: plt.clf()
2: plt.plot(predict, label='Predicted')
3: plt.plot(ans, label='Actual')
4: plt.legend()
5: plt.show()