fit

crowdkit.aggregation.embeddings.closest_to_average.ClosestToAverage.fit | Source code

fit(
    self,
    data: DataFrame,
    aggregated_embeddings: Optional[Series] = None,
    true_embeddings: Optional[Series] = None
)

Fits the model to the training data.

Parameters description

Parameters	Type	Description
`data`	DataFrame	The workers' outputs with their embeddings. The `pandas.DataFrame` data contains the `task`, `worker`, `output`, and `embedding` columns.
`aggregated_embeddings`	Optional[Series]	The task aggregated embeddings. The `pandas.Series` data is indexed by `task` and contains the corresponding aggregated embeddings.
`true_embeddings`	Optional[Series]	The embeddings of the true task responses. The `pandas.Series` data is indexed by `task` and contains the corresponding embeddings. The multiple true embeddings are not supported for a single task.

Crowd-Kit

Reference

Aggregation

Datasets

Learning

Metrics

Postprocessing