SKClassifier

class capymoa.base.SKClassifier

Bases: Classifier

A wrapper class for using scikit-learn classifiers in CapyMOA.

Some of scikit-learn’s classifiers that are compatible with online learning have been wrapped and tested already in CapyMOA (See capymoa.classifier).

However, if you want to use a scikit-learn classifier that has not been wrapped yet, you can use this class to wrap it yourself. This requires that the scikit-learn classifier implements the partial_fit and predict methods.

For example, the following code demonstrates how to use a scikit-learn classifier in CapyMOA:

>>> from sklearn.linear_model import SGDClassifier
>>> from capymoa.base import SKClassifier
>>> from capymoa.datasets import ElectricityTiny
>>> stream = ElectricityTiny()
>>> sklearner = SGDClassifier(random_state=1)
>>> learner = SKClassifier(sklearner, stream.schema)
>>> for _ in range(10):
...     instance = stream.next_instance()
...     prediction = learner.predict(instance)
...     print(f"True: {instance.y_index}, Predicted: {prediction}")
...     learner.train(instance)
True: 1, Predicted: None
True: 1, Predicted: 1
True: 1, Predicted: 1
True: 1, Predicted: 1
True: 0, Predicted: 1
True: 0, Predicted: 1
True: 0, Predicted: 0
True: 0, Predicted: 0
True: 0, Predicted: 0
True: 0, Predicted: 0

A word of caution: even compatible scikit-learn classifiers are not necessarily designed for online learning and might require some tweaking to work well in an online setting.

See also capymoa.base.SKRegressor for scikit-learn regressors.

__init__(

sklearner: ClassifierMixin,

schema: Schema = None,

random_seed: int = 1,

)

Construct a scikit-learn classifier wrapper.

Parameters:

• sklearner – A scikit-learn classifier object to wrap that must implements partial_fit and predict.

• schema – Describes the structure of the datastream.

• random_seed – Random seed for reproducibility.

Raises:

ValueError – If the scikit-learn algorithm does not implement partial_fit or predict.

sklearner: ClassifierMixin

The underlying scikit-learn object.

train(instance: LabeledInstance)

predict(instance: Instance)

predict_proba(instance: Instance)