The accuracy of the AD is only 0.81 and it has not changed the project. Is it because MSPCA is not used?

MAlessandrini-Univpm / rnn-eeg-ad

EEG-Based Alzheimer’s Disease Recognition Using Robust-PCA and LSTM Recurrent Neural Network

Creative Commons Attribution 4.0 International

12 stars 2 forks source link

The accuracy of the AD is only 0.81 and it has not changed the project. Is it because MSPCA is not used? #2

Open HanHenryOnlyFaceany opened 1 year ago

HanHenryOnlyFaceany commented 1 year ago

malessandrini commented 1 year ago

In the default script we artificially corrupt the data and then apply rpca to fix that. If you want to try the original data, set rnn.rpca=False and rnn.spikes=0. That said, some little random variations can still arise, due to different library versions, different GPU, etc.