tf.contrib.timeseries.saved_model_utils.cold_start_filter( signatures, session, features )
Perform filtering using an exported saved model.
Filtering refers to updating model state based on new observations. Predictions based on the returned model state will be conditioned on these observations.
Starts from the model's default/uninformed state.
MetaGraphDefprotocol buffer returned from
tf.saved_model.loader.load. Used to determine the names of Tensors to feed and fetch. Must be from the same model as
session: The session to use. The session's graph must be the one into which
tf.saved_model.loader.loadloaded the model.
features: A dictionary mapping keys to Numpy arrays, with several possible shapes (requires keys
FilteringFeatures.VALUES): Single example;
TIMESis a scalar and
VALUESis either a scalar or a vector of length [number of features]. Sequence;
TIMESis a vector of shape [series length],
VALUESeither has shape [series length] (univariate) or [series length x number of features] (multivariate). Batch of sequences;
TIMESis a vector of shape [batch size x series length],
VALUEShas shape [batch size x series length] or [batch size x series length x number of features]. In any case,
VALUESand any exogenous features must have their shapes prefixed by the shape of the value corresponding to the
A dictionary containing model state updated to account for the observations in
© 2018 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.