Solves a batch of isotonic regression problems.
tf.raw_ops.IsotonicRegression(
input,
output_dtype=tf.dtypes.float32
,
name=None
)
Args |
input
|
A Tensor . Must be one of the following types: float32 , float64 , int32 , uint8 , int16 , int8 , int64 , bfloat16 , uint16 , half , uint32 , uint64 .
A (batch_size, dim)-tensor holding a batch of inputs.
|
output_dtype
|
An optional tf.DType from: tf.half, tf.bfloat16, tf.float32, tf.float64 . Defaults to tf.float32 .
Dtype of output.
|
name
|
A name for the operation (optional).
|
Returns |
A tuple of Tensor objects (output, segments).
|
output
|
A Tensor of type output_dtype .
|
segments
|
A Tensor of type int32 .
|