Identifying mortality-related PM2.5 components and high-risk regions using machine learning

Chan Ju Kho

doi:10.1016/j.apr.2026.103107

Identifying mortality-related PM2.5 components and high-risk regions using machine learning

2026-06-15

0

PRE

AI

C

Chan Ju Kho

H

Hyemin Hwang

S

So Yeon Kim

J

Jeong-Hoo Park

E

Eunsun Jeong

H

Hee Young Kim

J

Jae Young Lee *

DOI：10.1016/j.apr.2026.103107

Original

Original request for help

Save

Abstract

En 中文

• Age-Standardized Mortality Rate (ASMR) were predicted using machine learning models • CatBoost showed the best performance under the 7-day lag condition • Among PM2.5 components, Nitrate and Arsenic were key features in ASMR prediction • SHapley Additive exPlanations identified thresholds for key components • High exceedance ratios for both components were observed in Seoul and Ansan

Identifying mortality-related PM2.5 components and high-risk regions using machine learning

Abstract

Journal

Organization

Cited Papers

Citing Papers