Industrial_ZSL

Source code of Industrial_ZSL on TE benchmark dataset.
The details of model can be found in
L. J. Feng, et al. Fault Description Based Attribute Transfer for Zero-Sample Industrial Fault Diagnosis, TII, 17(3), 1852 - 1862, 2021.

Fast execution in command line:

python3 TE_benchmark.py

Results Example:

==========================[test classes][3 ,6, 9]===================================
beginning...with feature extraction
loading data...
test classes: [3, 6, 9]
train classes: [0, 1, 2, 4, 5, 7, 8, 10, 11, 12, 13, 14].
SPCA extracting feature (takes lots of time)...
(5760, 400) (5760, 1) (5760, 20)
(1440, 400) (1440, 1) (1440, 20)
model training...
NB
accuracy: 0.6263888888888889
rf
accuracy: 0.5187777777777777

All rights reserved, citing the following papers are required for reference:

[1] L. J. Feng, et al. Fault Description Based Attribute Transfer for Zero-Sample Industrial Fault Diagnosis, TII, 17(3), 1852 - 1862, 2021.
[2] L. J. Feng, et al. Transfer Increment for Generalized Zero-Shot Learning, TNNLS, 32(6), 2506 - 2520, 2021.