skactiveml.stream.CognitiveDualQueryStrategy#

class skactiveml.stream.CognitiveDualQueryStrategy(budget_manager=None, budget=None, density_threshold=1, cognition_window_size=10, dist_func=None, dist_func_dict=None, random_state=None, force_full_budget=False)[source]#

Bases: SingleAnnotatorStreamQueryStrategy

This class is the base for the CognitiveDualQueryStrategy query strategy proposed in [1]. To use this strategy, refer to CognitiveDualQueryStrategyRan, CognitiveDualQueryStrategyRanVarUn, CognitiveDualQueryStrategyVarUn , and CognitiveDualQueryStrategyFixUn. The CognitiveDualQueryStrategy strategy is an extension to the uncertainty based query strategies proposed by Žliobaitė et al. [2] and follows the same idea as StreamDensityBasedAL [3] where queries for labels is only allowed if the local density around the corresponding instance is sufficiently high. The authors propose the use of a cognitive window that monitors the most representative samples within a data stream.

Parameters

budgetfloat, optional (default=None): The budget which models the budgeting constraint used in the stream-based active learning setting.
budget_managerBudgetManager, optional (default=None): The BudgetManager which models the budgeting constraint used in the stream-based active learning setting. if set to None, a default budget manager will be used that is defined in the class inheriting from CognitiveDualQueryStrategy. The budget manager will be initialized based on the following conditions:

If only a budget is given the default budget manager is initialized with the given budget. If only a budget manager is given use the budget manager. If both are not given the default budget manager with the default budget. If both are given and the budget differs from budgetmanager.budget a warning is thrown.
density_thresholdint, optional (default=1): Determines the local density factor size that needs to be reached in order to sample the candidate.
cognition_window_sizeint, optional (default=10): Determines the size of the cognition window
random_stateint, RandomState instance, optional (default=None): Controls the randomness of the estimator.
dist_funccallable, optional (default=None): The distance function used to calculate the distances within the local density window. If None use sklearn.metrics.pairwise.pairwise_distances
dist_func_dictdict, optional (default=None): Additional parameters for dist_func.
force_full_budgetbool, optional (default=False): If true, tries to utilize the full budget. The paper doesn’t update the budget manager if the locale density factor is 0

See also

budgetmanager.EstimatedBudgetZliobaite: BudgetManager implementing the base class for Zliobaite based budget managers
CognitiveDualQueryStrategyRan: CognitiveDualQueryStrategy using the RandomBudgetManager that is based on EstimatedBudgetZliobaite
CognitiveDualQueryStrategyFixUn: CognitiveDualQueryStrategy using the FixedUncertaintyBudgetManager that is based on EstimatedBudgetZliobaite
CognitiveDualQueryStrategyVarUn: VariableUncertaintyBudgetManager using the VariableUncertaintyBudgetManager that is based on EstimatedBudgetZliobaite
CognitiveDualQueryStrategyRanVarUn: CognitiveDualQueryStrategy using the RandomVariableUncertaintyBudgetManager that is based on EstimatedBudgetZliobaite

References

[1] Liu, S., Xue, S., Wu, J., Zhou, C., Yang, J., Li, Z., & Cao, J. (2021).: Online Active Learning for Drifting Data Streams. IEEE Transactions on Neural Networks and Learning Systems, 1-15.
[2] Žliobaitė, I., Bifet, A., Pfahringer, B., & Holmes, G. (2014). Active: Learning With Drifting Streaming Data. IEEE Transactions on Neural Networks and Learning Systems, 25(1), 27-39.
[3] Ienco, D., Pfahringer, B., & Zliobaitė, I. (2014). High density-focused: uncertainty sampling for active learning over evolving stream data. In BigMine 2014 (pp. 133-148).

Methods

`get_metadata_routing`()	Get metadata routing of this object.
`get_params`([deep])	Get parameters for this estimator.
`query`(candidates, clf[, X, y, ...])	Ask the query strategy which instances in candidates to acquire.
`set_params`(**params)	Set the parameters of this estimator.
`update`(candidates, queried_indices[, ...])	Updates the budget manager and the count for seen and queried instances

get_metadata_routing()#

Get metadata routing of this object.

Please check User Guide on how the routing mechanism works.

Returns

routingMetadataRequest: A MetadataRequest encapsulating routing information.

get_params(deep=True)#

Get parameters for this estimator.

Parameters

deepbool, default=True: If True, will return the parameters for this estimator and contained subobjects that are estimators.

Returns

paramsdict: Parameter names mapped to their values.

query(candidates, clf, X=None, y=None, sample_weight=None, fit_clf=False, return_utilities=False)[source]#

Ask the query strategy which instances in candidates to acquire.

Please note that, when the decisions from this function may differ from the final sampling, so the query strategy can be updated later with update(…) with the final sampling.

Parameters

candidates{array-like, sparse matrix} of shape
(n_samples, n_features): The instances which may be queried. Sparse matrices are accepted only if they are supported by the base query strategy.
clfSkactivemlClassifier: Model implementing the methods fit and predict_freq.
Xarray-like of shape (n_samples, n_features), optional
(default=None): Input samples used to fit the classifier.
yarray-like of shape (n_samples), optional (default=None): Labels of the input samples ‘X’. There may be missing labels.
sample_weightarray-like of shape (n_samples,), optional: Sample weights for X, used to fit the clf.
fit_clfbool, optional (default=False): If true, refit the classifier also requires X and y to be given.
return_utilitiesbool, optional (default=False): If true, also return the utilities based on the query strategy. The default is False.

Returns

queried_indicesndarray of shape (n_queried_instances,): The indices of instances in candidates which should be queried, with 0 <= n_queried_instances <= n_samples.
utilities: ndarray of shape (n_samples,), optional: The utilities based on the query strategy. Only provided if return_utilities is True.

set_params(**params)#

Set the parameters of this estimator.

The method works on simple estimators as well as on nested objects (such as Pipeline). The latter have parameters of the form <component>__<parameter> so that it’s possible to update each component of a nested object.

Parameters

**paramsdict: Estimator parameters.

Returns

selfestimator instance: Estimator instance.

update(candidates, queried_indices, budget_manager_param_dict=None)[source]#

Updates the budget manager and the count for seen and queried instances

Parameters

candidates{array-like, sparse matrix} of shape
(n_samples, n_features): The instances which could be queried. Sparse matrices are accepted only if they are supported by the base query strategy.
queried_indicesarray-like of shape (n_samples,): Indicates which instances from candidates have been queried.
budget_manager_param_dictkwargs, optional (default=None): Optional kwargs for budget_manager.

Returns

selfCognitiveDualQueryStrategy: The CognitiveDualQueryStrategy returns itself, after it is updated.

Examples using `skactiveml.stream.CognitiveDualQueryStrategy`#

Cognitive Dual-Query Strategy with Random Sampling

Cognitive Dual-Query Strategy with Fixed-Uncertainty

Cognitive Dual-Query Strategy with Variable-Uncertainty

Cognitive Dual-Query Strategy with Randomized-Variable-Uncertainty

skactiveml.stream.CognitiveDualQueryStrategy#

Examples using skactiveml.stream.CognitiveDualQueryStrategy#

Examples using `skactiveml.stream.CognitiveDualQueryStrategy`#