MiniRocket transforms input time series using a small, fixed set of convolutional kernels. MiniRocket uses PPV pooling to compute a single feature for each of the resulting feature maps (i.e., the proportion of positive values). The transformed features are used to train a linear classifier.
Dempster A, Schmidt DF, Webb GI (2020) MiniRocket: A Very Fast (Almost) Deterministic Transform for Time Series Classification arXiv:2012.08791
Import example data, MiniRocket, RidgeClassifierCV
(scikit-learn), and NumPy.
Note: MiniRocket and MiniRocketMultivariate are compiled by Numba on import. The compiled functions are cached, so this should only happen once (i.e., the first time you import MiniRocket or MiniRocketMultivariate).
# !pip install --upgrade numba
import numpy as np
from sklearn.linear_model import RidgeClassifierCV
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import StandardScaler
from aeon.datasets import load_arrow_head # univariate dataset
from aeon.datasets import load_basic_motions # multivariate dataset
from aeon.datasets import (
load_japanese_vowels, # multivariate dataset with unequal length
)
from aeon.transformations.collection.convolution_based import (
MiniRocket,
MiniRocketMultivariate,
MiniRocketMultivariateVariable,
)
For more details on the data set, see the univariate time series classification notebook.
Note: Input time series must be at least of length 9. Pad shorter time series using, e.g., PaddingTransformer
(aeon.transformers.panel.padder
).
X_train, y_train = load_arrow_head(split="train")
minirocket = MiniRocket() # by default, MiniRocket uses ~10_000 kernels
minirocket.fit(X_train)
X_train_transform = minirocket.transform(X_train)
# test shape of transformed training data -> (n_cases, 9_996)
X_train_transform.shape
(36, 9996)
We suggest using RidgeClassifierCV
(scikit-learn) for smaller datasets (fewer than ~10,000 training examples), and using logistic regression trained using stochastic gradient descent for larger datasets.
Note: For larger datasets, this means integrating MiniRocket with stochastic gradient descent such that the transform is performed per minibatch, not simply substituting RidgeClassifierCV
for, e.g., LogisticRegression
.
Note: While the input time-series of MiniRocket is unscaled, the output features of MiniRocket may need to be adjusted for following models. E.g. for RidgeClassifierCV
, we scale the features using the sklearn StandardScaler.
scaler = StandardScaler(with_mean=False)
classifier = RidgeClassifierCV(alphas=np.logspace(-3, 3, 10))
X_train_scaled_transform = scaler.fit_transform(X_train_transform)
classifier.fit(X_train_scaled_transform, y_train)
RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03]))In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03]))
X_test, y_test = load_arrow_head(split="test")
X_test_transform = minirocket.transform(X_test)
X_test_scaled_transform = scaler.transform(X_test_transform)
classifier.score(X_test_scaled_transform, y_test)
0.8742857142857143
Note: Input time series must be at least of length 9. Pad shorter time series using, e.g., PaddingTransformer
(aeon.transformers.panel.padder
).
X_train, y_train = load_basic_motions(split="train")
minirocket_multi = MiniRocketMultivariate()
minirocket_multi.fit(X_train)
X_train_transform = minirocket_multi.transform(X_train)
scaler = StandardScaler(with_mean=False)
X_train_scaled_transform = scaler.fit_transform(X_train_transform)
classifier = RidgeClassifierCV(alphas=np.logspace(-3, 3, 10))
classifier.fit(X_train_scaled_transform, y_train)
RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03]))In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03]))
X_test, y_test = load_basic_motions(split="test")
X_test_transform = minirocket_multi.transform(X_test)
X_test_scaled_transform = scaler.transform(X_test_transform)
classifier.score(X_test_scaled_transform, y_test)
1.0
# (above)
minirocket_pipeline = make_pipeline(
MiniRocket(),
StandardScaler(with_mean=False),
RidgeClassifierCV(alphas=np.logspace(-3, 3, 10)),
)
Note: Input time series must be at least of length 9. Pad shorter time series using, e.g., PaddingTransformer
(aeon.transformers.panel.padder
).
X_train, y_train = load_arrow_head(split="train")
# it is necessary to pass y_train to the pipeline
# y_train is not used for the transform, but it is used by the classifier
minirocket_pipeline.fit(X_train, y_train)
Pipeline(steps=[('minirocket', MiniRocket()), ('standardscaler', StandardScaler(with_mean=False)), ('ridgeclassifiercv', RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03])))])In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
Pipeline(steps=[('minirocket', MiniRocket()), ('standardscaler', StandardScaler(with_mean=False)), ('ridgeclassifiercv', RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03])))])
MiniRocket()
StandardScaler(with_mean=False)
RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03]))
X_test, y_test = load_arrow_head(split="test")
minirocket_pipeline.score(X_test, y_test)
0.8457142857142858
For a further pipeline, we use the extended version of MiniRocket, the MiniRocketMultivariateVariable
for variable / unequal length time series data. Following the code implementation of the original paper of miniRocket, we combine it with RidgeClassifierCV
in a sklearn pipeline. We can then use the pipeline like a self-contained classifier, with a single call to fit
, and without having to separately transform the data, etc.
Japanese vowels is a a UCI Archive dataset. 9 Japanese-male speakers were recorded saying the vowels ‘a’ and ‘e’. The raw recordings are preprocessed to get a 12-dimensional (multivariate) classification probem. The series lengths are between 7 and 29.
X_train_jv, y_train_jv = load_japanese_vowels(split="train")
# lets visualize the first three voice recordings with dimension 0-11
print("number of samples training: ", len(X_train_jv))
print("series length of recoding 0, dimension 5: ", X_train_jv[0][5].shape)
print("series length of recoding 1, dimension 0: ", X_train_jv[1][0].shape)
number of samples training: 270 series length of recoding 0, dimension 5: (20,) series length of recoding 1, dimension 0: (26,)
As before, we create a sklearn pipeline. MiniRocketMultivariateVariable requires a minimum series length of 9, where missing values are padded up to a length of 9, with the value "-10.0". Afterwards a scaler and a RidgeClassifierCV are added.
minirocket_mv_var_pipeline = make_pipeline(
MiniRocketMultivariateVariable(
pad_value_short_series=-10.0, random_state=42, max_dilations_per_kernel=16
),
StandardScaler(with_mean=False),
RidgeClassifierCV(alphas=np.logspace(-3, 3, 10)),
)
print(minirocket_mv_var_pipeline)
minirocket_mv_var_pipeline.fit(X_train_jv, y_train_jv)
Pipeline(steps=[('minirocketmultivariatevariable', MiniRocketMultivariateVariable(max_dilations_per_kernel=16, pad_value_short_series=-10.0, random_state=42)), ('standardscaler', StandardScaler(with_mean=False)), ('ridgeclassifiercv', RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03])))])
Pipeline(steps=[('minirocketmultivariatevariable', MiniRocketMultivariateVariable(max_dilations_per_kernel=16, pad_value_short_series=-10.0, random_state=42)), ('standardscaler', StandardScaler(with_mean=False)), ('ridgeclassifiercv', RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03])))])In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook.
Pipeline(steps=[('minirocketmultivariatevariable', MiniRocketMultivariateVariable(max_dilations_per_kernel=16, pad_value_short_series=-10.0, random_state=42)), ('standardscaler', StandardScaler(with_mean=False)), ('ridgeclassifiercv', RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03])))])
MiniRocketMultivariateVariable(max_dilations_per_kernel=16, pad_value_short_series=-10.0, random_state=42)
StandardScaler(with_mean=False)
RidgeClassifierCV(alphas=array([1.00000000e-03, 4.64158883e-03, 2.15443469e-02, 1.00000000e-01, 4.64158883e-01, 2.15443469e+00, 1.00000000e+01, 4.64158883e+01, 2.15443469e+02, 1.00000000e+03]))
Using the MiniRocketMultivariateVariable, we are able to process also process slightly larger input series than at train time. train max series length: 27, test max series length 29
X_test_jv, y_test_jv = load_japanese_vowels(split="test")
minirocket_mv_var_pipeline.score(X_test_jv, y_test_jv)
0.9945945945945946