AutoTune Function about emggo HOT 1 CLOSED

The purpose of the AutoTune function is to find parameters from the data and the adjusted onsets/offsets. It is supposed to be the converse operation of AutoFind which takes data and an estimated number of onset/offset pairs. AutoTune can be applied in a situation such as:

"We have 100 data files each with a large number of muscle activation intervals. We do not want to manually label all the files. We are struggling to find a suitable set of parameters by hit and trial. The AutoTune solves this problem. Take one file at random. Manually label it from scratch (or use AutoFind first to get an initial set of onsets/offsets). Auto tune parameters from this file and apply these parameters to the rest of the 99 files."

The problem is of course the question: "How well these parameters from 1 file will behave for the remaining 99 files?" I do not know. The current version is based on the above idea but not fully tested yet. This is something I am still working on and it needs additional systematic investigation. This is why it was not part of the publication (https://doi.org/10.1016/j.jelekin.2019.06.010).

from emggo.