Back to models
Regressor

TapNetRegressor

Time series attentional prototype network (TapNet), as described in [1].

Quickstart

python
from sktime.regression.deep_learning.tapnet import TapNetRegressor

estimator = TapNetRegressor(n_epochs=2000, batch_size=16, dropout=0.5, filter_sizes=(256, 256, 128), kernel_size=(8, 5, 3), dilation=1, layers=(500, 300), use_rp=True, activation='linear', activation_hidden='leaky_relu', rp_params=(-1, 3), use_bias=True, use_att=True, use_lstm=True, use_cnn=True, random_state=None, padding='same', loss='mean_squared_error', optimizer=None, metrics=None, callbacks=None, verbose=False, lstm_dropout=0.8)

Parameters(20)

filter_sizesarray of int, default = (256, 256, 128)
sets the kernel size argument for each convolutional block. Controls number of convolutional filters and number of neurons in attention dense layers.
kernel_sizearray of int, default = (8, 5, 3)
controls the size of the convolutional kernels
layersarray of int, default = (500, 300)
size of dense layers
n_epochsint, default = 2000
number of epochs to train the model
batch_sizeint, default = 16
number of samples per update
callbackslist of keras.callbacks.Callback, optional (default=None)
List of Keras callbacks to apply during model training.
dropoutfloat, default = 0.5
dropout rate, in the range [0, 1)
lstm_dropoutfloat, default = 0.8
dropout rate for the LSTM layer, in the range [0, 1)
dilationint, default = 1
dilation value
activationstr, default = “linear”

activation function for the last output layer List of available activation functions: https://keras.io/api/layers/activations/

activation_hiddenstr, default = “leaky_relu”

activation function for the hidden layers List of available activation functions: https://keras.io/api/layers/activations/

lossstr, default = “mean_squared_error”
loss function for the classifier
optimizerstr or None, default = “Adam(lr=0.01)”
gradient updating function for the classifier
use_biasbool, default = True
whether to use bias in the output dense layer
use_rpbool, default = True
whether to use random projections
use_attbool, default = True
whether to use self attention
use_lstmbool, default = True
whether to use an LSTM layer
use_cnnbool, default = True
whether to use a CNN layer
verbosebool, default = False
whether to output extra information
random_stateint or None, default = None
seed for random

References

  1. [1 ] Zhang et al. Tapnet: Multivariate time series classification with attentional prototypical network, Proceedings of the AAAI Conference on Artificial Intelligence 34(4), 6845-6852, 2020