The impact of neglecting feature scaling in k-means clustering

Chantha Wongoutong · 2024

Read The impact of neglecting feature scaling in k-means clustering by Chantha Wongoutong (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11623793.txt.

Description

Chantha Wongoutong, writing alone in 2024, tests what happens when data are not scaled before a popular grouping method is run. The method, k-means, sorts data into clusters, and scaling is a step that is often skipped.

Five scaling methods were compared with raw data on sets whose measures shared a unit and on sets whose measures did not. When units differed, scaling first gave better accuracy and other scores than raw data. When units matched, scaling changed little. Two methods, z-score and percentile scaling, did best on the mixed unit sets. The author says scaling should be a firm step before grouping such data.

Catalog record

CREATORChantha Wongoutong
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE36 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC11623793

Browse the full library · About the Tower of Babel