Page 1 of 1

Variable name of the normalizer output

Posted: 09 Jan 2012, 12:14
by vladi
Hallo,

I am trying to develop a algorithm which should take the output of the normalizier which is a value between -1 and 1 as I understand and put this in a buffer for 5 trials. Afterwards I would like to calculate the true positive rate from the values in the buffer and reset the offset value of the classiefer. And than for the next 5 trials and so on. So I am trying to find the output variable in the code of the classifier but I am not certain which variable it is?? I would be really greatfull if anybody could help me out.

Re: Variable name of the normalizer output

Posted: 10 Jan 2012, 10:58
by mellinger
Hi,
the output of the normalizier which is a value between -1 and 1
the output of the normalizer is not between -1 and 1. Rather, it has a variance of 1, and a mean of 0. This means that most of its values are indeed between -1 and 1, but there is no guarantee that this is always the case.
Afterwards I would like to calculate the true positive rate from the values in the buffer and reset the offset value of the classiefer.
If by "classifier" you mean the LinearClassifier filter, then this has no offset. Rather, the offset is dynamically adapted by the Normalizer. So it seems what you actually want to do is to achieve a certain behavior of the Normalizer. From what you write, it appears that the Normalizer can already do what you want. Please make sure you read the Normalizer's wiki page at:
http://www.bci2000.org/wiki/index.php/U ... Normalizer

Regards,
Juergen