search

MAMEM / eeg-processing-toolbox

Matlab code for proccesing EEG signals.
http://www.mamem.eu/
Apache License 2.0
76 stars 38 forks source link

Use of canonical correlation analysis for Dataset II #62

Open ShivaNaghsh opened 6 years ago

ShivaNaghsh commented 6 years ago

Hi, Could you please guide me to learn how to use these codes for canonical correlation analysis? Actually, I am going to work on Dataset II, and I can currently split the sessions into the trials, but I don't know how to apply canonical correlation method to them... Thank you very much for any help you could provide.

ShivaNaghsh commented 5 years ago

@liarosge I saw you have answered all questions, I would appreciate it if you help me solve my problem with use of CCA method using your Matlab codes.

vangelis2015 commented 5 years ago

Hello, First it is needed to familiarize yourself with the toolbox. Start by looking the example files: exampleCombiCCA.m, exampleITCCA.m and exampleEpoc.m. In order to choose the desired CCA based method you need to define the correct class (for example: the line "classif = eegtoolkit.classification.ITCCA;" defines the ITCCA method as the method of analysis)

ShivaNaghsh commented 5 years ago

Thank you for your response @vangelis2015. I need to work with a simple form of cca; however, these examples seem to use complex versions of cca, so I am wondering if it is enough to use "canoncorr" which is the matlab's function for cca method and has been used in cca.m file too??

vangelis2015 commented 5 years ago

In order to be consisted with the structure and the overall scheme of the toolbox, i recommend to modify the ITCCA classifier (it can be found on folder eegtoolkit/classification). You can define your own templates by treating carefully the class variable: individualTemplates. Alternatively, you can modify exampleEpoc.m, which use the classical CCA, to met your requirements.