Back to the main page.
Bug 3377 - lambda parameter in beamformer tutorial exercise 6
| Status | NEW |
| Reported | 2017-11-27 14:07:00 +0100 |
| Modified | 2017-12-05 15:43:12 +0100 |
| Product: | FieldTrip |
| Component: | documentation |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P5 normal |
| Assigned to: | KrisitijanArmeni |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Jan-Mathijs Schoffelen - 2017-11-27 14:07:33 +0100
Currently, exercise 6 in the beamformer tutorial is didactically sub-optimal, because it asks for the exploration of different amounts of regularization, which don't seem to be making much of a difference. Every time when running this tutorial we have to apologetically say that based on the example regularization does not seem to be making much of a difference. I was just playing with some other data, and it seems that the value needs to be cranked up substantially, in order to appreciate the smoothing effect of this parameter. Thus, I suggest to use values such as '100%', or even '500%' to make the point. http://www.fieldtriptoolbox.org/tutorial/beamformer