Using the Kth Nearest Neighbor Clustering Procedure to Determine the Number of Subpopulations (Classic Reprint) - Hardcover

M. Anthony Wong

9780484220866: Using the Kth Nearest Neighbor Clustering Procedure to Determine the Number of Subpopulations (Classic Reprint)

Hardcover

ISBN 10: 0484220861 ISBN 13: 9780484220866

Publisher: Forgotten Books, 2018

View all copies of this ISBN edition

0 Used

2 New

From US$ 32.25

Discover how to uncover hidden subpopulations in data with a practical, study‑ready method.

This guide explains a kth nearest neighbor clustering approach to count and test the number of modes in a univariate density, helping you see when a population is unimodal, bimodal, or more complex.

The book walks you through a clear procedure: estimate a density with a chosen k, use hierarchical clustering to identify modes, and apply a bootstrap test to assess significance. It emphasizes when the method works best and where small or uneven subpopulations can affect results, all in accessible, data‑driven terms.

How the diagnostic plot signals the number of modes and possible substructures in a data set.
How to use the bootstrap‑based p‑value to decide if observed multimodality is statistically significant.
Practical notes on sample size effects, computation time, and limitations in higher dimensions.
Illustrative examples from real data to show interpretation in context.

Ideal for readers of applied statistics, data scientists, and researchers who want a concrete, nonparametric tool to explore population structure without heavy modeling assumptions. It offers a tested workflow you can adapt to your own data.

"synopsis" may belong to another edition of this title.