I'm trying to make a prediction of an eeg signal

but i'm having some difficulties

for example

i tried to use the RandomForest classifier with a simple db2 transformation of my discrete signal

but the results have not been satisfactory, 67% of accuracy

after i changed to Adaboost

and i have almost the same thing

I think that i need to find a better way to precisely distinguish eeg time series patterns, or try something to improve some parameters of the algorithms that i'm using

@amoeba As it stands, this does not really fit SE standards. Either we need to convert the latest thread as a comment, or you will need to use a dedicated chat room (or create a new session of our (dead) Journal Club :-).

@chl: The discussion seems to be over, so there is no sense in creating a chat room now. Do you think something NEEDS to be done with the existing thread? If so, either mods need to move the whole thread into the comments of the first David's answer splitting his second answer into 2-3 comments; or (if such splitting is not possible), edit the second answer into the first, and move the comments.

But, honestly, I would prefer if as many people as have already commented here on how to format the thread, would comment on what they think about the discussion itself :)

Hi is CV still online? I'm having trouble accessing the site (other SE sites are fine). Browser says connection refused.

Hi, is someone available for a really small and quick question (I just need to know the technical name of something)