我试图在张量流中设计一个简单的lstm。我想将一系列数据分类为1到10个类。
我有 10个时间戳和数据X.我现在只采用一个序列，所以我的批量大小= 1。 在每个时期，都会生成一个新序列。例如，X是像这样的numpy数组 -
X [[ 2.52413028 2.49449348 2.46520466 2.43625973 2.40765466 2.37938545
2.35144815 2.32383888 2.29655379 2.26958905]]
为了使它适合于lstm输入，我首先转换为张量然后重新整形（batch_size，sequence_lenght，输入维度） -
X= np.array([amplitude * np.exp(-t / tau)])
print 'X', X
#Sorting out the input
train_input = X
train_input = tf.convert_to_tensor(train_input)
train_input = tf.reshape(train_input,[1,10,1])
print 'ti', train_input
对于输出，我在1到10的类范围内生成一个热编码标签。
#------------sorting out the output
train_output= [int(math.ceil(tau/resolution))]
train_output= one_hot(train_output, num_labels=10)
print 'label', train_output
train_output = tf.convert_to_tensor(train_output)
>>label [[ 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]]
然后我为张量流图创建了占位符，制作了lstm单元并给出了权重和偏差 -
data = tf.placeholder(tf.float32, shape= [batch_size,len(t),1])
target = tf.placeholder(tf.float32, shape = [batch_size, num_classes])
cell = tf.nn.rnn_cell.LSTMCell(num_hidden)
output, state = rnn.dynamic_rnn(cell, data, dtype=tf.float32)
weight = tf.Variable(tf.random_normal([batch_size, num_classes, 1])),
bias = tf.Variable(tf.random_normal([num_classes]))
#training
prediction = tf.nn.softmax(tf.matmul(output,weight) + bias)
cross_entropy = -tf.reduce_sum(target * tf.log(prediction))
optimizer = tf.train.AdamOptimizer()
minimize = optimizer.minimize(cross_entropy)
到目前为止，我已经编写了代码并且在训练步骤中出错了。是否与输入形状有关？这是追溯---
追踪（最近一次呼叫最后一次）：
File "/home/raisa/PycharmProjects/RNN_test1/test3.py", line 66, in <module>
prediction = tf.nn.softmax(tf.matmul(output,weight) + bias)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/math_ops.py", line 1036, in matmul
name=name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/gen_math_ops.py", line 911, in _mat_mul
transpose_b=transpose_b, name=name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/op_def_library.py", line 655, in apply_op
op_def=op_def)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/ops.py", line 2156, in create_op
set_shapes_for_outputs(ret)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/ops.py", line 1612, in set_shapes_for_outputs
shapes = shape_func(op)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/common_shapes.py", line 81, in matmul_shape
a_shape = op.inputs[0].get_shape().with_rank(2)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/tensor_shape.py", line 625, in with_rank
raise ValueError("Shape %s must have rank %d" % (self, rank))
ValueError: Shape (1, 10, 5) must have rank 2
答案 0 :(得分：2)
查看代码时，您的rnn输出的维度应为public class GridManager : GridManagerBase<IGridPoint>
{
public GridManager()
{
GridTarget = new GridPoint();
SelectedTarget = new TargetObject(1, 2);
}
public void DoWork()
{
var result = GridTarget.HasVehicle; // works
var result2 = ((ITargetObject)SelectedTarget).SomeValue; // works with cast
SelectedTarget.DoSomeCommonWork();
}
}
public class GridManagerB : GridManagerBase<IGridPoint>
{
public GridManagerB()
{
GridTarget = new GridPoint();
SelectedTarget = new TargetObjectB(1, string.Empty);
}
public void DoWork()
{
var result = GridTarget.HasVehicle; // works
var result2 = ((ITargetObjectB)SelectedTarget).SomeOtherValue; // works with cast
SelectedTarget.DoSomeCommonWork();
}
}
public abstract class GridManagerBase<TGridPoint> : IGridManagerBase<TGridPoint>
where TGridPoint : class, IGridPointBase
{
public TGridPoint GridTarget { get; protected set; }
public ITargetObjectBase SelectedTarget { get; protected set; }
//public IVehicleObjectBase SelectedVehicle { get; protected set; }
}
public interface IGridManagerBase<TGridPoint>
{
TGridPoint GridTarget { get; }
ITargetObjectBase SelectedTarget { get; }
}
public interface IGridPointBase
{
int Row { get; }
}
public interface IGridPoint : IGridPointBase
{
bool HasVehicle { get; }
}
public interface ITargetObjectBase
{
int Row { get; set; }
void DoSomeCommonWork();
}
public interface ITargetObject : ITargetObjectBase
{
int SomeValue { get; }
}
public interface ITargetObjectB : ITargetObjectBase
{
string SomeOtherValue { get; }
}
public abstract class TargetObjectBase : ITargetObjectBase
{
public int Row { get; set; }
public TargetObjectBase(int row)
{
Row = row;
}
public void DoSomeCommonWork()
{
Console.WriteLine(Row.ToString());
}
}
public class TargetObject : TargetObjectBase, ITargetObject
{
public int SomeValue { get; private set; }
public TargetObject(int row, int some)
: base(row)
{
SomeValue = some;
}
}
public class TargetObjectB : TargetObjectBase, ITargetObjectB
{
public string SomeOtherValue { get; private set; }
public TargetObjectB(int row, string other)
: base(row)
{
SomeOtherValue = other;
}
}
public class GridPoint : IGridPoint
{
public int Row { get; set; }
public bool HasVehicle { get; set; }
}
，而w的维度为batch_size x 1 x num_hidden，但您希望这两者的乘法为batch_size x num_classes x 1。
您可以尝试batcH_size x num_classes和output = tf.reshape(output, [batch_size, num_hidden])并告诉我这是怎么回事？
答案 1 :(得分：1)
如果您使用的是TF＆gt; = 1.0，则可以利用tf.contrib.rnn库和OutputProjectionWrapper将完全连接的图层添加到RNN的输出中。类似的东西：
# Network definition.
cell = tf.contrib.rnn.LSTMCell(num_hidden)
cell = tf.contrib.rnn.OutputProjectionWrapper(cell, num_classes) # adds an output FC layer for you
output, state = tf.nn.dynamic_rnn(cell, data, dtype=tf.float32)
# Training.
cross_entropy = tf.nn.softmax_cross_entropy_with_logits(logits=output, labels=targets)
cross_entropy = tf.reduce_sum(cross_entropy)
optimizer = tf.train.AdamOptimizer()
minimize = optimizer.minimize(cross_entropy)
注意我使用softmax_cross_entropy_with_logits代替您使用prediction操作并手动计算交叉熵。它应该更有效和更强大。
OutputProjectionWrapper基本上做同样的事情，但它可能有助于减轻一些麻烦。