some problems

Specific discussion regarding applications.
Post Reply
Jessica
Junior
Posts: 2
Joined: Thu Mar 6, 2008 14:39

some problems

Post by Jessica »

Dear Norbert,

Thank you for the generous help. Now I have tried the CRP tool to our experimental data recorded from the hippocampal neurons. And I have some problems in the applications.

Firstly, which method has been used to choose the threshold in CRP? Although many methods have been proposed in the paper 'Recurrence plots for the analysis of complex systems', I'm still confused because of the moment change in spike trains.

Secondly, in this case the choice of ε depends strongly on the considered system understudy, could the method of unthreshold be used? How to choose the parameter in this method?

Lastly, is embedding is neccessary? The embeded dimensions calculated by various methods are quite high, which may lead to spurious results. Sometimes, no appropriate embeded parameters could be gained, such as burst.

PS: It's my first time to apply nonlinear science to our experimental data, please excuse for so many silly questions.

Sincerely,
Jessica
Post Reply