<div dir="ltr"><span style="font-size:12.8000001907349px">Hi to all specialist.</span><div style="font-size:12.8000001907349px">I am a post graduate student and my thesis is around recognizing and classifying brain signals.</div><div style="font-size:12.8000001907349px">I've been having problem splitting the signals into the trials that I can train a artificial neuron network.</div><div style="font-size:12.8000001907349px">I've been using a dataset from BCI Competition 2008 "Graz dataset B" and I load the signal via this command :</div><div style="font-size:12.8000001907349px"><br></div><div style="font-size:12.8000001907349px"><font face="monospace, monospace" size="4">[s , h] = sload('B0101T.gdf')</font></div><div style="font-size:12.8000001907349px"><font face="monospace, monospace" size="4"><br></font></div><div style="font-size:12.8000001907349px"><font face="arial, helvetica, sans-serif">and it loads signals as s and header as h into the workspace. s includes 6 columns refer to C3 , C4 and Cz. h includes 3 sub-variables as h.EVENT.POS (position of occurrence) h.EVENT.TYP ( type ) and h.EVENT.DUR ( duration).</font></div><div style="font-size:12.8000001907349px"><span style="font-family:arial,helvetica,sans-serif">I need you guidance on how to extract trials and the use these trials to make a feature vector to train the ann????</span></div><div style="font-size:12.8000001907349px"><span style="font-family:arial,helvetica,sans-serif">Another question is that should i use the 3 channels or one of them is sufficient?</span></div><div style="font-size:12.8000001907349px"><span style="font-family:arial,helvetica,sans-serif">I appreciate your response.</span></div><div style="font-size:12.8000001907349px"><span style="font-family:arial,helvetica,sans-serif">Thanks a lot. </span></div>
</div>