ewannurse / a-generalizable-bci-using-machine-learning-for-feature-discovery

Repository of data and source code for publication 'A Generalizable BCI using Machine Learning for Feature Discovery'

A-Generalizable-BCI-using-Machine-Learning-for-Feature-Discovery

Repository of data and source code for publication 'A Generalizable BCI using Machine Learning for Feature Discovery'

Full publication can be found at http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0131328

Please cite this article if using this data set.

Full citation: Nurse ES, Karoly PJ, Grayden DB, Freestone DR (2015) A Generalizable Brain-Computer Interface (BCI) Using Machine Learning for Feature Discovery. PLoS ONE 10(6): e0131328. doi: 10.1371/journal.pone.0131328

Bibtex: @article{10.1371/journal.pone.0131328, author = {Nurse, Ewan S. AND Karoly, Philippa J. AND Grayden, David B. AND Freestone, Dean R.}, journal = {PLoS ONE}, publisher = {Public Library of Science}, title = {A Generalizable Brain-Computer Interface (BCI) Using Machine Learning for Feature Discovery}, year = {2015}, month = {06}, volume = {10}, url = {http://dx.doi.org/10.1371%2Fjournal.pone.0131328}, pages = {e0131328}, number = {6}, doi = {10.1371/journal.pone.0131328} }

a-generalizable-bci-using-machine-learning-for-feature-discovery's Issues

Code Execution

I am trying to run your program on the given data sets. I would like to clarify below things:

• Have you used only data set present at (data/) ?
• I am not able to figure out which of your program files read the above data folder (from below list)
  Fitness_Test.m
  Mutate.m
  Neural_Network_Classifier.m
  Pop_2_Binary.m
  Pop_2_Dec.m
  Roulette_Cross.m
• Which program file is the main() program file (Like in Python) where all execution starts
• On your paper , section <3.2 Methord> : says 4646 pixel images formed for caffe execution. Would you please help me how you got this and also where u coded on your program file.
  Below is the text
  " Each signal consisted of a 2D matrix with
  dimensions 46�100 (Figure 5), with the vertical axis corre-
  sponding to the 46 measurement channels (corresponding to
  46 scalp electrodes) and the horizontal axis corresponding to
  100 datapoints across the time domain. These signals were
  rst ltered to remove outlier spikes and then resized to a set
  of 4646 pixel images for conversion into a Cafe-compatible
  dataset."