tf.feature_column.weighted_categorical_column

TensorFlow 2 version

View source on GitHub

Applies weight values to a CategoricalColumn.

tf.feature_column.weighted_categorical_column(
    categorical_column, weight_feature_key, dtype=tf.dtypes.float32
)

Use this when each of your sparse inputs has both an ID and a value. For example, if you're representing text documents as a collection of word frequencies, you can provide 2 parallel sparse input features ('terms' and 'frequencies' below).

Example:

Input tf.Example objects:

[
  features {
    feature {
      key: "terms"
      value {bytes_list {value: "very" value: "model"} }
    }
    feature {
      key: "frequencies"
      value {float_list {value: 0.3 value: 0.1} }
    }
  },
  features {
    feature {
      key: "terms"
      value {bytes_list {value: "when" value: "course" value: "human"} }
    }
    feature {
      key: "frequencies"
      value {float_list {value: 0.4 value: 0.1 value: 0.2} }
    }
  }
]

categorical_column = categorical_column_with_hash_bucket(
    column_name='terms', hash_bucket_size=1000)
weighted_column = weighted_categorical_column(
    categorical_column=categorical_column, weight_feature_key='frequencies')
columns = [weighted_column, ...]
features = tf.io.parse_example(..., features=make_parse_example_spec(columns))
linear_prediction, _, _ = linear_model(features, columns)

This assumes the input dictionary contains a SparseTensor for key 'terms', and a SparseTensor for key 'frequencies'. These 2 tensors must have the same indices and dense shape.