Weighted KNN
Hello to everybody,
I have a small question about weighting in KNN classifiers. Just to see if I understood well.
Does "Triangular" weighting that neighbors closest to the "unknown" sample are weighted more in class assigment, and that weight decrease linerly with distance?
Does "Rectangular" (default) means that all neighbors have the same weight?
I have also seen that there are other weighting patters, some with strange names. Do you have some information about them?
Thanks in advance.
Comments
Hi,
You can visualize the weighting scheme via a plot in the k-nn analysis. This will plot the weight as a function of the relative distance of the neighbors.
See also https://epub.ub.uni-muenchen.de/1769/1/paper_399.pdf about more exact info about the weights.
Koen
OK. Thank you very much.
Leonardo