Question: Suppose we have a classification problem where the training data are tuples (yi , xi) with xi = [xi1, ..., xiD] R D. Suppose we
Suppose we have a classification problem where the training data are tuples (yi , xi) with xi = [xi1, ..., xiD] R D. Suppose we re-scale the feature vectors x i = [xi1/s1, ..., xiD/sD] using different scale factors sd > 0 for different dimensions d. Is the KNN classification function using standard Euclidean distance invariant to this re-scaling? That is, will it always be the case that fKNN (xi) = fKNN (x i )? Explain your
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
