BoostedTreesAggregateStats

public final class BoostedTreesAggregateStats

Aggregates the summary of accumulated stats for the batch.

The summary stats contains gradients and hessians accumulated for each node, feature dimension id and bucket.

Public Methods

Output<Float>
asOutput()
Returns the symbolic handle of a tensor.
static BoostedTreesAggregateStats
create(Scope scope, Operand<Integer> nodeIds, Operand<Float> gradients, Operand<Float> hessians, Operand<Integer> feature, Long maxSplits, Long numBuckets)
Factory method to create a class wrapping a new BoostedTreesAggregateStats operation.
Output<Float>
statsSummary()
output Rank 4 Tensor (shape=[splits, feature_dimension, buckets, logits_dimension + hessian_dimension]) containing accumulated stats for each node, feature dimension and bucket.

Inherited Methods

Public Methods

public Output<Float> asOutput ()

Returns the symbolic handle of a tensor.

Inputs to TensorFlow operations are outputs of another TensorFlow operation. This method is used to obtain a symbolic handle that represents the computation of the input.

public static BoostedTreesAggregateStats create (Scope scope, Operand<Integer> nodeIds, Operand<Float> gradients, Operand<Float> hessians, Operand<Integer> feature, Long maxSplits, Long numBuckets)

Factory method to create a class wrapping a new BoostedTreesAggregateStats operation.

Parameters
scope current scope
nodeIds int32; Rank 1 Tensor containing node ids for each example, shape [batch_size].
gradients float32; Rank 2 Tensor (shape=[batch_size, logits_dimension]) with gradients for each example.
hessians float32; Rank 2 Tensor (shape=[batch_size, hessian_dimension]) with hessians for each example.
feature int32; Rank 2 feature Tensors (shape=[batch_size, feature_dimension]).
maxSplits int; the maximum number of splits possible in the whole tree.
numBuckets int; equals to the maximum possible value of bucketized feature.
Returns
  • a new instance of BoostedTreesAggregateStats

public Output<Float> statsSummary ()

output Rank 4 Tensor (shape=[splits, feature_dimension, buckets, logits_dimension + hessian_dimension]) containing accumulated stats for each node, feature dimension and bucket.