Open ManTouCIBR opened 1 year ago
22000 because there are 22 channels recorded during 4 seconds (from seconds 3 to 7) with a sampling frequency of 250Hz. for more information: https://github.com/EdgarMoyete/EEG-Classification-with-CNN https://www.bbci.de/competition/iv/desc_2a.pdf https://www.bbci.de/competition/iv/ https://ieeexplore.ieee.org/document/9629958
22000 because there are 22 channels recorded during 4 seconds (from seconds 3 to 7) with a sampling frequency of 250Hz. for more information: https://github.com/EdgarMoyete/EEG-Classification-with-CNN https://www.bbci.de/competition/iv/desc_2a.pdf https://www.bbci.de/competition/iv/ https://ieeexplore.ieee.org/document/9629958
Thanks again for your reply, the link and reply you gave are very helpful for my study!
Hello author, through the CSV data of BCI 2A you shared, I realized a neural network and got good classification results. However, when I explained the network to my teacher, the teacher pointed out a point that I had ignored: the information problem carried by the data itself. The data of 9 people in CSV shared are all the data volume of (288,22000). Here, I would like to ask you: 288 is understandable. There are four experimental paradigms recorded 72 times each, and 288 times in total. How does 22,000 understand that? What channels of data is he recording? Is it the data of BCI 3S or 1+ 3S with 1s extra rest time?If you have time to answer, it will be very helpful to me! Anyway, thank you again!