我正在用Keras构建LSTM预测器。我的输入数组是历史价格数据。我将数据划分为window_size个块,以便提前预测prediction length个块。我的数据是4246个浮点数的列表。我将数据分成4055个数组,每个数组的长度为168,以便预测前面的24个单位。

这给我一个尺寸为x_train(4055,168)设置。然后,我缩放数据并尝试拟合数据,但遇到尺寸错误。

df = pd.DataFrame(data)
print(f"Len of df: {len(df)}")
min_max_scaler = MinMaxScaler()
H = 24

window_size = 7*H
num_pred_blocks = len(df)-window_size-H+1

x_train = []
y_train = []
for i in range(num_pred_blocks):
    x_train_block = df['C'][i:(i + window_size)]
    x_train.append(x_train_block)
    y_train_block = df['C'][(i + window_size):(i + window_size + H)]
    y_train.append(y_train_block)

LEN = int(len(x_train)*window_size)
x_train = min_max_scaler.fit_transform(x_train)
batch_size = 1

def build_model():
    model = Sequential()
    model.add(LSTM(input_shape=(window_size,batch_size),
                   return_sequences=True,
                   units=num_pred_blocks))
    model.add(TimeDistributed(Dense(H)))
    model.add(Activation("linear"))
    model.compile(loss="mse", optimizer="rmsprop")
    return model

num_epochs = epochs
model= build_model()
model.fit(x_train, y_train, batch_size = batch_size, epochs = 50)


这样返回的错误就是这样。

ValueError: Error when checking model target: the list of Numpy arrays that you are passing to your model is not the size the model expected. Expected to see 1 array(s), but instead got the following list of 4055 arrays: [array([[0.00630006],


我无法正确分割吗?正确加载?单元数应该与预测块数不同吗?感谢您的帮助。谢谢。

编辑

将它们转换为Numpy数组的建议是正确的,但MinMixScalar()返回一个numpy数组。我将阵列重塑为适当的尺寸,但是现在我的计算机出现CUDA内存错误。我认为问题已经解决。谢谢。

df = pd.DataFrame(data)
min_max_scaler = MinMaxScaler()
H = prediction_length

window_size = 7*H
num_pred_blocks = len(df)-window_size-H+1

x_train = []
y_train = []
for i in range(num_pred_blocks):
    x_train_block = df['C'][i:(i + window_size)].values
    x_train.append(x_train_block)
    y_train_block = df['C'][(i + window_size):(i + window_size + H)].values
    y_train.append(y_train_block)

x_train = min_max_scaler.fit_transform(x_train)
y_train = min_max_scaler.fit_transform(y_train)
x_train = np.reshape(x_train, (len(x_train), 1, window_size))
y_train = np.reshape(y_train, (len(y_train), 1, H))
batch_size = 1

def build_model():
    model = Sequential()
    model.add(LSTM(batch_input_shape=(batch_size, 1, window_size),
                   return_sequences=True,
                   units=100))
    model.add(TimeDistributed(Dense(H)))
    model.add(Activation("linear"))
    model.compile(loss="mse", optimizer="rmsprop")
    return model

num_epochs = epochs
model = build_model()
model.fit(x_train, y_train, batch_size = batch_size, epochs = 50)

最佳答案

我认为您没有在模型中通过批次大小。

input_shape=(window_size,batch_size)是数据维度。正确,但是您应该使用input_shape=(window_size, 1)

如果要使用批处理,则必须添加另一个尺寸,例如LSTM(n_neurons, batch_input_shape=(n_batch, X.shape[1], X.shape[2]))(引用自Keras)

在您的情况下:

def build_model():
    model = Sequential()
    model.add(LSTM(input_shape=(batch_size, 1, window_size),
                   return_sequences=True,
                   units=num_pred_blocks))
    model.add(TimeDistributed(Dense(H)))
    model.add(Activation("linear"))
    model.compile(loss="mse", optimizer="rmsprop")
    return model


您还需要使用np.shape更改数据的维,它应该是(batch_dimdata_dim_1data_dim_2)。我使用numpy,所以numpy.reshape()可以工作。

首先,您的数据应按行排列,因此对于每一行,您都应具有(1, 168)的形状,然后添加批处理维度,它将为(batch_n, 1, 168)

希望能有所帮助。

关于python - Keras LSTM输入/输出尺寸,我们在Stack Overflow上找到一个类似的问题:https://stackoverflow.com/questions/59020754/

10-15 22:45