A program to recognize and reward our most engaged community members
what is the meaning of “ k-th nearest neighbors” ？--operator “Detect Outlier (Distances)”，
Should k-th nearest distance be used or the mean of k nearest distance be used?
Hi AP - When they say the Kth value can be a number between 1 and a million, they just mean the algorithm will consider aspects of the one nearest neighbor to a known point, or two nearest neighbors, or three, or four or up to the one million nearest neighbors to a known point, depending on the number you select for k.
Once the algorithm finds the k nearest neighbors, it might use the mean of some attribute of all those points as the best guess at what the same attribute on your original point is, or some other calculation, but the calculation uses the k nearest points.
Hope this helps.
