אריה קנטורוביץ

אקדמי בכיר

Nearest-Neighbor Methods

A Modern Perspective

Aryeh Kontorovich, Samory Kpotufe

This chapter aims at providing an overview of various modern approaches to learning with nearest neighbors in general metric spaces. We provide the necessary background and then proceed to cover classification, regression—with sufficient detail and literature pointers to yield practical insights into how various configuration and pre-processing choices, e.g., metric, the number of neighbors, data subsampling, and compression, affect learning and computational performance.

שפת פרסום אנגלית
דפים 75-92
סטטוס פרסום פורסם - 01.01.2023

ASJC Scopus subject areas

General Computer Science
General Mathematics
גישה למסמך
10.1007/978-3-031-24628-9_5
קבצים וקישורים אחרים
Link to publication in Scopus