. Author manuscript; available in PMC: 2023 Feb 1.

Published in final edited form as: IEEE Trans Knowl Data Eng. 2020 Apr 22;34(2):531–543. doi: 10.1109/tkde.2020.2989405

TABLE 6:

The architecture of the infused, poor model used by CHEER, Direct, KD and AT for knowledge infusion in PTBDB, which includes a total 45.0k parameters.

Layer	Type	Hyper-parameters	Activation
1	Split	n_seg=10
2	Convolution1D	n_filter=32, kernel_size=16, stride=2	ReLU
3	Convolution1D	n_filter=32, kernel_size=16, stride=2	ReLU
4	Convolution1D	n_filter=32, kernel_size=16, stride=2	ReLU
5	AveragePooling1D
6	LSTM	hidden_units=32	ReLU
7	PositionAttention
8	Dense	hidden_units=n_classes	Linear
9	Softmax