An adaptive nearest neighbor rule for classification
share
We introduce a variant of the k-nearest neighbor classifier in which k is chosen adaptively for each query, rather than supplied as a parameter. The choice of k depends on properties of each neighborhood, and therefore may significantly vary between different points. (For example, the algorithm will use larger k for predicting the labels of points in noisy regions.) We provide theory and experiments that demonstrate that the algorithm performs comparably to, and sometimes better than, k-NN with an optimal choice of k. In particular, we derive bounds on the convergence rates of our classifier that depend on a local quantity we call the `advantage' which is significantly weaker than the Lipschitz conditions used in previous convergence rate proofs. These generalization bounds hinge on a variant of the seminal Uniform Convergence Theorem due to Vapnik and Chervonenkis; this variant concerns conditional probabilities and may be of independent interest.
READ FULL TEXT