Simple CNN for knotted x unknotted recognition
The model is trained on downsampled knots_AF dataset where sequences with length < 500 are padded with 'X's.
The architecture is PENGUINN with modified input.
Model versions
(names in format cnn_{NUMBER_OF_EPOCHS}epochs_{TEST_ACCURACY}.h5)
Trained on v1 version of the dataset (it contained duplicites):
Trained on v2 version of the dataset (some unknotted labels were incorrect):
Trained on the latest version of the dataset:
cnn_10epochs_latest.h5- accuracy on test:~95.17 %