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.
|
