actsnclass.DataBase.make_query

DataBase.make_query(strategy='UncSampling', batch=1, screen=False, queryable=False, query_thre=1.0) → list

Identify new object to be added to the training sample.

Parameters:
  • strategy (str (optional)) – Strategy used to choose the most informative object. Current implementation accepts ‘UncSampling’ and ‘RandomSampling’, ‘UncSamplingEntropy’, ‘UncSamplingLeastConfident’, ‘UncSamplingMargin’, ‘QBDMI’, ‘QBDEntropy’, . Default is UncSampling.
  • batch (int (optional)) – Number of objects to be chosen in each batch query. Default is 1.
  • queryable (bool (optional)) – If True, consider only queryable objects. Default is False.
  • query_thre (float (optional)) – Percentile threshold where a query is considered worth it. Default is 1 (no limit).
  • screen (bool (optional)) – If true, display on screen information about the displacement in order and classificaion probability due to constraints on queryable sample.
Returns:

query_indx – List of indexes identifying the objects to be queried in decreasing order of importance. If strategy==’RandomSampling’ the order is irrelevant.
Return type:

list