feat: Implement HybridKNearestNeighbors

docs/additional_information/changelog.rst

@@ -27,6 +27,7 @@ Added
 - Implement and integrate type validation into ``dtaianomaly``.
 - Added marks to the test: ``slow`` and ``numba``.
 - Integrated numpydocs documentation formatting rules.
+- Implement ``HybridKNearestNeighbors`` anomaly detector.
 
 Changed
 ^^^^^^^

docs/additional_information/contributing.rst

@@ -268,4 +268,4 @@ Documentation
 | |check_box| Has a citation to the relevant paper(s) been added in the class documentation, and the bibtex added to ``docs/bibliography.bib``?
 | |check_box| Did you add the component in tha API-documentation in ``docs/api/``?
 | |check_box| If you added a :py:class:`~dtaianomaly.data.LazyDataLoader`, did you update `data/README.rst <https://github.com/ML-KULeuven/dtaianomaly/blob/main/data/README.rst>`_?
-
+| |check_box| Have you updated the changelog?

docs/api/anomaly_detection.rst

@@ -72,6 +72,7 @@ Neural methods
 
     AutoEncoder
     ConvolutionalNeuralNetwork
+    HybridKNearestNeighbors
     LongShortTermMemoryNetwork
     MultilayerPerceptron
     Transformer

docs/bibliography.bib

@@ -525,3 +525,17 @@ @InProceedings{perini2021quantifying
 }
 
 
+@article{song2017hybrid,
+    author = {Song, Hongchao and Jiang, Zhuqing and Men, Aidong and Yang, Bo},
+    title = {A Hybrid Semi-Supervised Anomaly Detection Model for High-Dimensional Data},
+    journal = {Computational Intelligence and Neuroscience},
+    volume = {2017},
+    number = {1},
+    pages = {8501683},
+    doi = {https://doi.org/10.1155/2017/8501683},
+    year = {2017}
+}
+
+
+
+

dtaianomaly/anomaly_detection/_BasePyODAnomalyDetector.py

@@ -6,6 +6,7 @@
 from dtaianomaly.anomaly_detection._BaseDetector import BaseDetector, Supervision
 from dtaianomaly.type_validation import IntegerAttribute, WindowSizeAttribute
 from dtaianomaly.windowing import (
+    WINDOW_SIZE_TYPE,
     compute_window_size,
     reverse_sliding_window,
     sliding_window,
@@ -42,7 +43,7 @@ class BasePyODAnomalyDetector(BaseDetector, abc.ABC):
         The PyOD anomaly detector
     """
 
-    window_size: int | str
+    window_size: WINDOW_SIZE_TYPE
     stride: int
     kwargs: dict
     window_size_: int
@@ -53,7 +54,7 @@ class BasePyODAnomalyDetector(BaseDetector, abc.ABC):
         "stride": IntegerAttribute(1),
     }
 
-    def __init__(self, window_size: int | str, stride: int = 1, **kwargs):
+    def __init__(self, window_size: WINDOW_SIZE_TYPE, stride: int = 1, **kwargs):
         super().__init__(self._supervision())
         self.window_size = window_size
         self.stride = stride