CNNClassifierSKTimeWrapper
Python template · sinapsis_sktime.templates.sktime_classifiers
Overview
Template for SkTime classification model training and inference for the deep_learning module.
CNNClassifier — inherited documentation
Time Convolutional Neural Network (CNN), as described in [1]_.
Zhao et al. 2017 uses sigmoid activation in the hidden layers. To obtain same behaviour as Zhao et al. 2017, set activation_hidden to "sigmoid".
Adapted from the implementation from Fawaz et al. https://github.com/hfawaz/dl-4-tsc/blob/master/classifiers/cnn.py
Parameters ---------- n_epochs : int, default = 2000 the number of epochs to train the model batch_size : int, default = 16 the number of samples per gradient update. kernel_size : int, default = 7 the length of the 1D convolution window avg_pool_size : int, default = 3 size of the average pooling windows n_conv_layers : int, default = 2 the number of convolutional plus average pooling layers callbacks : list of keras.callbacks, default = None verbose : boolean, default = False whether to output extra information loss : string, default="categorical_crossentropy" fit parameter for the keras model metrics : list of strings, default=["accuracy"] random_state : int or None, default=None Seed for random number generation. activation : string or tf callable, default="softmax" Activation function used in the output layer. List of available activation functions: https://keras.io/api/layers/activations/ activation_hidden : string or tf callable, default="sigmoid" Activation function used in the hidden layers. List of available activation functions: https://keras.io/api/layers/activations/ Default value of activation_hidden will change to "relu" in version '0.42.0'. use_bias : boolean, default = True whether the layer uses a bias vector. optimizer : keras.optimizers object, default = Adam(lr=0.01) specify the optimizer and the learning rate to be used. filter_sizes : array of shape (n_conv_layers) default = [6, 12] padding : string, default = "auto" Controls padding logic for the convolutional layers, i.e. whether ``'valid'`` and ``'same'`` are passed to the ``Conv1D`` layer. - "auto": as per original implementation, ``"same"`` is passed if ``input_shape[0] < 60`` in the input layer, and ``"valid"`` otherwise. - "valid", "same", and other values are passed directly to ``Conv1D``
References ---------- .. [1] Zhao et al., Convolutional neural networks for time series classification, Journal of Systems Engineering and Electronics, 28(1):2017.
Examples -------- >>> from sktime.classification.deep_learning.cnn import CNNClassifier >>> from sktime.datasets import load_unit_test >>> X_train, y_train = load_unit_test(split="train") >>> X_test, y_test = load_unit_test(split="test") >>> cnn = CNNClassifier(n_epochs=20, batch_size=4) # doctest: +SKIP >>> cnn.fit(X_train, y_train) # doctest: +SKIP CNNClassifier(...)
Capabilities
Capabilities: not setInstallation
Install the published package using pip or uv.
pip install sinapsis-sktime==1.0.0uv pip install sinapsis-sktime==1.0.0uv add sinapsis-sktime==1.0.0Configuration
Effective fields from the inspector schema, including nested Pydantic model fields.
Optional
Default: null
Optional
Default: "/sinapsis-hub/private/platform/sinapsis-hub/playground/publish"
Path where the trained model will be saved. cnnclassifier_init n_epochs <class 'int'>: default: 2000 batch_size <class 'int'>: default: 16 kernel_size <class 'int'>: default: 7 avg_pool_size <class 'int'>: default: 3… Read full field description
Path where the trained model will be saved. cnnclassifier_init n_epochs <class 'int'>: default: 2000 batch_size <class 'int'>: default: 16 kernel_size <class 'int'>: default: 7 avg_pool_size <class 'int'>: default: 3 n_conv_layers <class 'int'>: default: 2 callbacks <class 'NoneType'>: default: None verbose <class 'bool'>: default: False loss <class 'str'>: default: categorical_crossentropy metrics <class 'NoneType'>: default: None random_state typing.Optional[int]: default: None activation <class 'str'>: default: softmax activation_hidden <class 'str'>: default: changing_from_sigmoid_to_relu_in_0.42.0 use_bias <class 'bool'>: default: True optimizer <class 'NoneType'>: default: None filter_sizes <class 'NoneType'>: default: None padding <class 'str'>: default: auto
Required
Default: null
Required
Default: null
Optional
Default: "categorical_crossentropy"
Optional
Default: null
Optional
Default: "auto"
Optional
Default: false
Optional
Default: 2000
Optional
Default: true
Optional
Default: null
Optional
Default: null
Optional
Default: "softmax"
Optional
Default: 16
Optional
Default: 7
Optional
Default: null
Optional
Default: null
Optional
Default: 3
Optional
Default: 2
Optional
Default: "changing_from_sigmoid_to_relu_in_0.42.0"
Configuration example
Generated from documented defaults. Replace any placeholders and complete required nested settings before running.
templates:
- template_name: "CNNClassifierSKTimeWrapper"
class_name: "CNNClassifierSKTimeWrapper"
attributes:
root_dir: "/sinapsis-hub/private/platform/sinapsis-hub/playground/publish"
model_save_path: "<required>"
cnnclassifier_init:
loss: "categorical_crossentropy"
metrics: null
padding: "auto"
verbose: false
n_epochs: 2000
use_bias: true
callbacks: null
optimizer: null
activation: "softmax"
batch_size: 16
kernel_size: 7
filter_sizes: null
random_state: null
avg_pool_size: 3
n_conv_layers: 2
activation_hidden: "changing_from_sigmoid_to_relu_in_0.42.0"
Inheritance
sinapsis_sktime.templates.sktime_classifiers.SKTimeDeepLearningClassifiersEnvironment variables
Environment definitions belong to their canonical Python modules.