nnsa.feature_extraction.brainagemodel.pretrained package

Subpackages

nnsa.feature_extraction.brainagemodel.pretrained.models package
- Module contents

Submodules

nnsa.feature_extraction.brainagemodel.pretrained.convert module

nnsa.feature_extraction.brainagemodel.pretrained.prenorm module

Functions:

`get_norm_pars`()
`pre_normalize`(eeg_in_volts)

nnsa.feature_extraction.brainagemodel.pretrained.prenorm.get_norm_pars()[source]

nnsa.feature_extraction.brainagemodel.pretrained.prenorm.pre_normalize(eeg_in_volts)[source]

nnsa.feature_extraction.brainagemodel.pretrained.pretrainedsincmodel module

Classes:

PretrainedSincModel(CH[, verbose])

To load all ensembled trained Sinc models.

class nnsa.feature_extraction.brainagemodel.pretrained.pretrainedsincmodel.PretrainedSincModel(CH, verbose=True)[source]

Bases: EnsembleModels

To load all ensembled trained Sinc models. Note that it is not a keras model, but it has a predict funtion with similar inputs/outputs. It also has a predict_recording which also applies the recording-level aggregation and then model ensembling aggregation according to the paper.

Parameters:

CH: the number of eeg channels {1, 2, 4, or 8} verbose: if True, it shows the loading progress; otherwise, it is silent.