I want to share this repo where we evaluate ACC performance between KAN and MLP in four scenarios: RAW Multiclass, RAW Binary, CSP Multiclass, and CSP Binary.
Which report?
I could only find a notebook, inside "results", which shows approximately the same results, except for one which shows a bit more overfitting on KAN.
I want to share this repo where we evaluate ACC performance between KAN and MLP in four scenarios: RAW Multiclass, RAW Binary, CSP Multiclass, and CSP Binary.
Draw your conclusions (see ours in the report).
Best!
https://github.com/JARS29/KAN_MLP_MI-BCI