tf_agents.bandits.multi_objective.multi_objective_scalarizer.LinearScalarizer

View source on GitHub

Scalarizes multple objectives by a linear combination.

Inherits From: Scalarizer

tf_agents.bandits.multi_objective.multi_objective_scalarizer.LinearScalarizer(
    weights: Sequence[tf_agents.bandits.multi_objective.multi_objective_scalarizer.ScalarFloat]
)

Args
weights	A Sequence of weights for linearly combining the objectives.

Raises
TypeError	if not isinstance(weights, Sequence).

Methods

call

View source

call(
    multi_objectives: tf.Tensor
) -> tf.Tensor

Implementation of scalarization logic by subclasses.

__call__

View source

__call__(
    multi_objectives: tf.Tensor
) -> tf.Tensor

Returns a single reward by scalarizing multiple objectives.

Args
multi_objectives	A Tensor of shape [batch_size, number_of_objectives], where each column represents an objective.

Returns: A Tensor of shape [batch_size] representing scalarized rewards.

Raises
ValueError	if multi_objectives.shape.rank != 2.
ValueError	if multi_objectives.shape.dims[1] != self._num_of_objectives.