Add description and (incomplete) doctest for TEDA-CDD

river/drift/teda_cdd.py

@@ -6,7 +6,61 @@
 
 
 class TEDACDD(DriftDetector):
-    """ """
+    r"""Concept Drift Detector based on Typically and Eccentrically Data Analytics (TEDA-CDD)
+
+    Concept Drift Detector based on Typically and Eccentrically Data Analytics (TEDA-CDD) [^1] is a concept drift detector
+    based on TEDA [^2], a framework for data analytic leveraging on typicality and eccentricity. It employs two models in
+    monitoring the data stream in order to keep the information of a previous concept whereas monitoring the emergence
+    of a new concept. The models are considered to represent to distinct concepts when the intersection of data samples
+    are significantly low, described by the Jaccard Index (JI).
+
+    TEDA-CDD has four essential components: reference data model, evolving data model, detection metric, and reset
+    strategy. The reference model uses the classical definition of TEDA to represent the concept known by the classifier,
+    disregarding any atypical data sample, whereas the evolving model uses an adaptation to describe the current features
+    state.
+
+    The model indicates a concept drift when the reference and evolving models are sufficiently distinct. In this context,
+    a detection strategy based on the JI, considering the concept models as representations of sets, is proposed. Moreover,
+    when a concept drift is detected, TEDA-CDD is reset to start monitoring the current concept; and, to avoid a cold
+    restart, the information on the evolving model is used to update the reference model while a new evolving model is
+    created from scratch.
+
+    Parameters
+    ----------
+    m
+        Sensitivity parameter [^2]. A higher value of `m` makes it difficult to encounter atypical data samples, whereas
+        lower values causes more data samples to be considered atypical.
+    alpha
+        The forgetting factor, affecting the evolving model as a sensitivity parameter.
+        The higher the value of `alpha`, the more important is a data sample to the current concept and the evolving
+        model is more sensible to noise.
+    jaccard_threshold
+        The Jaccard Threshold (JT). A concept drift occurs if the jaccard index is lower than this given threshold.
+        This threshold defines a limit of similarity between similar and dissimilar and the higher the value of JT, the
+        more sensible to divergence is the CDD.
+
+    Examples
+    ----------
+    >>> import random
+    >>> from river import drift
+
+    >>> rng = random.Random(12345)
+    >>> teda_cdd = TEDACDD(m=1, alpha=0.95, jaccard_threshold=0.95)
+
+    >>> # Simulate a data stream composed by two data distributions
+    >>> data_stream = rng.choices([0, 1], k=1000) + rng.choices(range(4, 8), k=1000)
+
+    >>> # Update drift detector and verify if change is detected
+    >>> for i, val in enumerate(data_stream):
+    ...     teda_cdd.update(val)
+
+    References
+    ----------
+    [^1]: Y. T. P. Nunes, L. A. Guedes. 2024. Concept Drift Detection Based on Typicality and Eccentricity. IEEE Access
+    12 (2024), pp. 13795-13808. doi: 10.1109/ACCESS.2024.3355959.
+    [^2]: P. Angelov. 2014. Anomaly detection based on eccentricity analysis. 2014 IEEE Symposium on Evolving and
+    Autonomous Learning Systems (EALS), Orlando, FL, USA, pp. 1-8. doi: 10.1109/EALS.2014.7009497.
+    """
 
     def __init__(self, m=3.0, alpha=0.95, jaccard_threshold=0.8):
         super().__init__()