Developer Guide for Intel® Data Analytics Acceleration Library 2018 Update 3

Batch Processing

Layer Input

The backward one-dimensional max pooling layer accepts the input described below. Pass the Input ID as a parameter to the methods that provide input for your algorithm. For more details, see Algorithms.

Input ID

Input

inputGradient

Pointer to tensor G of size l1 x ... x lp that stores the input gradient computed on the preceding layer. This input can be an object of any class derived from Tensor.

inputFromForward

Collection of data needed for the backward one-dimensional max pooling layer.

Element ID

Element

auxSelectedIndices

Tensor T of size l1 x ... x lp that stores indices of maximum elements.

auxInputDimensions

NumericTable of size 1 x p that stores the sizes of the dimensions of input data tensor X: n1, n2, …, np.

Layer Parameters

For common parameters of neural network layers, see Common Parameters.

In addition to the common parameters, the backward one-dimensional max pooling layer has the following parameters:

Parameter

Default Value

Description

algorithmFPType

float

The floating-point type that the algorithm uses for intermediate computations. Can be float or double.

method

defaultDense

Performance-oriented computation method, the only method supported by the layer.

kernelSize

KernelSize(2)

Data structure representing the size of the one-dimensional subtensor from which the maximum element is selected.

stride

Stride(2)

Data structure representing the interval on which the subtensors for max pooling are selected.

padding

Padding(0)

Data structure representing the number of data elements to implicitly add to each side of the one-dimensional subtensor along which max pooling is performed.

index

Index(p-1)

Index k of the dimension along which max pooling is performed.

Layer Output

The backward one-dimensional max pooling layer calculates the result described below. Pass the Result ID as a parameter to the methods that access the results of your algorithm. For more details, see Algorithms.

Result ID

Result

gradient

Pointer to tensor Z of size n1 x ... x np that stores the result of the backward one-dimensional max pooling layer. This input can be an object of any class derived from Tensor.

Examples

C++: max_pool1d_layer_dense_batch.cpp

Java*: MaxPool1DLayerDenseBatch.java

Python*: max_pool1d_layer_dense_batch.py

See Also