On an Example in the Cover-Hart Paper on Nearest Neighbor Classification
Year: 2025
Type: preprint
Abstract: In this note, we revisit the Example in Section VII of the classical Cover-Hart paper on the asymptotic error of the nearest neighbor classification rule. The expression given there for the expected classification error, though asymptotically correct, is slightly off at finite sample sizes. Here, we provide a detailed new derivation of the correct formula.
Source: HAL
Author: Braga-Neto, Ulisses
Institution: Texas A&M University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7119383363
