Hello,
I am working on my Master’s thesis, and need some information on Emotiv online classification. I am unable to use Openvibe for online classification, therefore I can only use Emotiv to evaluate the online performance. I am running an offline analysis and need it to match the Emotiv procedure as closely as possible, to properly compare the offline results from openvibe. I need some specifics about Emotiv’s classification procedure. Please help me out, your black-box methods aren’t very supportive of research.
1: What Machine learning algorithms are you using? SVM, LDA, MLP, ANN?
2: What Processing do you do to extract features? I have only seen talk of Power Spectral Density for power bands, and Independent Component Analysis for facial expressions. Do you use others such as, Common Spatial Patterns?
3: Feature selection..Exactly what features, and how many do you use for your classifier?
You may also email me at Jameskuffuor@gmail.com in case you cannot reply publicly.