Weighted k Nearest Neighbours


When you’re working with a kNN classifier, you have to normalize your data to ensure all your parameters are equally considered when computing the nearest neighbours.
Is there a classifier where you normalize all your data, but then multiply some parameters by a factor a > 0 (so the range for min/max normalization is [0, a]) ? That way, more importance would be put on a particular parameter compared to others.

Thanks !