Scikit-Learn compatible HMM and DTW based sequence machine learning algorithms in Python.
KNNMixin._dtw1d
when independent=True
by @eonu in https://github.com/eonu/sequentia/pull/251
Full Changelog: https://github.com/eonu/sequentia/compare/v2.0.1...v2.0.2
KNNClassifier.predict_log_proba
by @eonu in https://github.com/eonu/sequentia/pull/247
Full Changelog: https://github.com/eonu/sequentia/compare/v2.0.0...v2.0.1
scikit-learn
compatibility + general refactor by @eonu in https://github.com/eonu/sequentia/pull/241
README.md
logo URL by @eonu in https://github.com/eonu/sequentia/pull/243
Full Changelog: https://github.com/eonu/sequentia/compare/v1.1.1...v2.0.0
max_nbytes=None
to fix read-only buffer source array error in joblib.Parallel
(see https://github.com/scikit-learn/scikit-learn/issues/7981). (#235)sequentia.preprocessing
module with sklearn.preprocessing
compatibility. (#234)sequentia.pipeline
module for sklearn.pipeline
compatibility. (#234)CategoricalHMM
and GaussianMixtureHMM
parameter defaults for params
/init_params
being modified. (#231)CategoricalHMM
and GaussianMixtureHMM
unfreeze()
calling super().freeze()
instead of super().unfreeze()
. (#231)_KNNMixin
when weighting=None
. (#231)preprocessing
module (temporarily until design is finalized). (#226)datasets
module for sample datasets. (#226)datasets.load_random_sequences
for generating an arbitrarily sized dataset of sequences. (#216)DeepGRU
and classifier.rnn
module. (#215)sequentia.datasets
module. (#214)return_scores
argument to KNNClassifier.predict()
to return class scores. (#213)self
in fit()
functions. (#213)hmmlearn
v0.2.7. (#201)HMMClassifier
structure to match KNNClassifier
. (#200)'uniform'
KNNClassifier
weighting option. (#192)KNNClassifier
label scoring bug - thanks @manisci. (#187)digits.npz
as package data in setup.py
. (#221)CONTRIBUTING.md
CI instructions. (#219)datasets
module. (#217)tslearn
as a core dependency. (#216)torchaudio
, torchvision
and torchfsdd
dependencies. (#214)play_audio
helper. (#214)README.md
and documentation. (#202)Jinja2
dependency for RTD. (#188)
KNNClassifier
has a major bug in all versions prior to and including v0.12.1 resulting in inaccurate predictions (see #186).
GMMHMM
andHMMClassifier
have a major bug in all versions prior to and including v0.12.1 as a result of two bugs in theGMMHMM
class inhmmlearn
versions before v0.2.7 (see #193).⚠️ Please use version v0.13.0 or later.
requirements.py
due to import error. (#182)