Classifier

CanonicalIntervalForest

Canonical Interval Forest Classifier (CIF).

Implementation of the interval based forest making use of the catch22 feature set on randomly selected intervals described in Middlehurst et al. (2020). [1]

Overview: Input “n” series with “d” dimensions of length “m”. For each tree:

Sample n_intervals intervals of random position and length
Subsample att_subsample_size catch22 or summary statistic attributes randomly
Randomly select dimension for each interval
Calculate attributes for each interval, concatenate to form new data set
Build decision tree on new data set

ensemble the trees with averaged probability estimates

Schnellstart

python

from sktime.classification.interval_based import CanonicalIntervalForest

estimator = CanonicalIntervalForest(n_estimators=200, n_intervals=None, att_subsample_size=8, min_interval=3, max_interval=None, base_estimator='CIT', n_jobs=1, random_state=None)

Parameter(8)

n_estimatorsint, default=200: Number of estimators to build for the ensemble.
n_intervalsint or None, default=None: Number of intervals to extract per tree, if None extracts (sqrt(series_length) * sqrt(n_dims)) intervals.
att_subsample_sizeint, default=8: Number of catch22 or summary statistic attributes to subsample per tree.
min_intervalint, default=3: Minimum length of an interval.
max_intervalint or None, default=None: Maximum length of an interval, if None set to (series_length / 2).
base_estimatorsklearn classifier or str, default=”CIT”.: Base estimator for the ensemble, can be supplied a sklearn BaseEstimator or a string for predefined classifiers. Possible strings:
"CIT", uses the sktime ContinuousIntervalTree, an implementation of the original tree used with embedded attribute processing for faster predictions.
"DTC" uses the sklearn DecisionTreeClassifier(criterion="entropy").
n_jobsint, default=1: The number of jobs to run in parallel for both fit and predict. -1 means using all processors.
random_stateint or None, default=None: Seed for random number generation.

Beispiele

>>> from sktime.classification.interval_based import CanonicalIntervalForest
>>> from sktime.datasets import load_unit_test
>>> X_train, y_train = load_unit_test (split = "train", return_X_y = True)
>>> X_test, y_test = load_unit_test (split = "test", return_X_y = True)
>>> clf = CanonicalIntervalForest (
... n_estimators = 3, n_intervals = 2, att_subsample_size = 2
... )
>>> clf. fit (X_train, y_train) CanonicalIntervalForest(
... )
>>> y_pred = clf. predict (X_test)

Referenzen

[1]

Matthew Middlehurst and James Large and Anthony Bagnall. “The Canonical Interval Forest (CIF) Classifier for Time Series Classification.” IEEE International Conference on Big Data 2020