Hello,
A point in KNN doesn't normally take itself into account if I understood well, so why does it say it returns it's own label?
Thanks
During training, yes one point is its own neighbour. So the training error with 1-NN is always 0.
2019 Ex13
Hello,
A point in KNN doesn't normally take itself into account if I understood well, so why does it say it returns it's own label?
Thanks
During training, yes one point is its own neighbour. So the training error with 1-NN is always 0.
Add comment