循环神经网络 (RNN) 旨在学习序列数据。正如您所猜测的，它们绝对可以将多个特征作为输入！Keras 的 RNN 采用时间步T和特征F的 2D 输入（ T，F ） （我在这里忽略了批量维度）。

但是，您并不总是需要或想要中间时间步长t = 1, 2 ... ( T - 1)。因此，Keras 灵活地支持这两种模式。要让它输出所有T个时间步，请在构建时传递return_sequences=True给您的 RNN（例如，LSTM或GRU）。如果您只想要最后一个时间步t = T，则使用return_sequences=False（如果您不传递return_sequences给构造函数，这是默认设置）。

以下是这两种模式的示例。

示例 1：学习序列

这是一个训练 LSTM（RNN 类型）的快速示例，它可以保持整个序列。在这个例子中，每个输入数据点有 2 个时间步长，每个时间步长有 3 个特征；输出数据有 2 个时间步长（因为return_sequences=True），每个时间步长有 4 个数据点（因为这是我传递给的大小LSTM）。

import keras.layers as L
import keras.models as M

import numpy

# The inputs to the model.
# We will create two data points, just for the example.
data_x = numpy.array([
    # Datapoint 1
    [
        # Input features at timestep 1
        [1, 2, 3],
        # Input features at timestep 2
        [4, 5, 6]
    ],
    # Datapoint 2
    [
        # Features at timestep 1
        [7, 8, 9],
        # Features at timestep 2
        [10, 11, 12]
    ]
])

# The desired model outputs.
# We will create two data points, just for the example.
data_y = numpy.array([
    # Datapoint 1
    [
        # Target features at timestep 1
        [101, 102, 103, 104],
        # Target features at timestep 2
        [105, 106, 107, 108]
    ],
    # Datapoint 2
    [
        # Target features at timestep 1
        [201, 202, 203, 204],
        # Target features at timestep 2
        [205, 206, 207, 208]
    ]
])

# Each input data point has 2 timesteps, each with 3 features.
# So the input shape (excluding batch_size) is (2, 3), which
# matches the shape of each data point in data_x above.
model_input = L.Input(shape=(2, 3))

# This RNN will return timesteps with 4 features each.
# Because return_sequences=True, it will output 2 timesteps, each
# with 4 features. So the output shape (excluding batch size) is
# (2, 4), which matches the shape of each data point in data_y above.
model_output = L.LSTM(4, return_sequences=True)(model_input)

# Create the model.
model = M.Model(input=model_input, output=model_output)

# You need to pick appropriate loss/optimizers for your problem.
# I'm just using these to make the example compile.
model.compile('sgd', 'mean_squared_error')

# Train
model.fit(data_x, data_y)

示例 2：学习最后一个时间步

另一方面，如果你想训练一个只输出序列中最后一个时间步长的 LSTM，那么你需要设置return_sequences=False（或者只是将它从构造函数中完全删除，因为False这是默认设置）。然后您的输出数据（data_y在上面的示例中）需要重新排列，因为您只需要提供最后一个时间步。所以在第二个例子中，每个输入数据点仍然有 2 个时间步长，每个时间步长有 3 个特征。然而，输出数据只是每个数据点的单个向量，因为我们已将所有内容扁平化为单个时间步长。但是，这些输出向量中的每一个仍然具有 4 个特征（因为这是我传递给的大小LSTM）。

import keras.layers as L
import keras.models as M

import numpy

# The inputs to the model.
# We will create two data points, just for the example.
data_x = numpy.array([
    # Datapoint 1
    [
        # Input features at timestep 1
        [1, 2, 3],
        # Input features at timestep 2
        [4, 5, 6]
    ],
    # Datapoint 2
    [
        # Features at timestep 1
        [7, 8, 9],
        # Features at timestep 2
        [10, 11, 12]
    ]
])

# The desired model outputs.
# We will create two data points, just for the example.
data_y = numpy.array([
    # Datapoint 1
    # Target features at timestep 2
    [105, 106, 107, 108],
    # Datapoint 2
    # Target features at timestep 2
    [205, 206, 207, 208]
])

# Each input data point has 2 timesteps, each with 3 features.
# So the input shape (excluding batch_size) is (2, 3), which
# matches the shape of each data point in data_x above.
model_input = L.Input(shape=(2, 3))

# This RNN will return timesteps with 4 features each.
# Because return_sequences=False, it will output 2 timesteps, each
# with 4 features. So the output shape (excluding batch size) is
# (2, 4), which matches the shape of each data point in data_y above.
model_output = L.LSTM(4, return_sequences=False)(model_input)

# Create the model.
model = M.Model(input=model_input, output=model_output)

# You need to pick appropriate loss/optimizers for your problem.
# I'm just using these to make the example compile.
model.compile('sgd', 'mean_squared_error')

# Train
model.fit(data_x, data_y)