I am working on miBCI using the dataset above mentioned in python. For preprocessing of the same I've used a butterworth filter the result of which is then used for Independent Component Analysis.
This is the code using sklearn.
ica = FastICA()
X_raw1= ica.fit_transform(X_raw1.T)
I have also attached the plot of the resultant data obtained. Please help me verify if the code works well and what could be further implied by the plots.
Thank You in advance.
How many channels you are using?
And what is your purpose of using ICA? Noise removal?
Dataset is having readings from 3 channels each of 9 seconds sampled at 128hz frequency and 140 such trials.
And I am using ICA for noise removal.
There seems to be an apparent difference between the signals you obtained from left and right movements. Now try to find a feature which will show the apparent difference between them. You may start with entropies: in Matlab you can have various entropies calculated by "wentropy" function e.g. log energy entropy, norm entropy etc.
Best wishes
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python