Skip to main navigation Skip to search Skip to main content

Exploring the Determinants of Well-Being: Insights from SHAP and ICE Analyses

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This study applies Automated Machine Learning (AutoML) and Explainable Artificial Intelligence (XAI) techniques to identify important factors influencing individual well-being. Using data from the “Human Information Database” and a well-being survey, a CatBoost regression model was selected as the best model. SHapley Additive Explanations (SHAP) and Individual Conditional Expectation (ICE) were then used to analyze the impact of various factors on well-being, both overall and at a type-specific level. Additionally, SHAP-based clustering was performed to group individuals with similar well-being drivers. A marketing-focused individual-level analysis was also conducted. These results provide actionable insights for developing personalized strategies to enhance well-being, offering practical applications through data-driven analysis.

Original languageEnglish
Title of host publicationInformation Management - 11th International Conference, ICIM 2025, Revised Selected Papers
EditorsShuliang Li
PublisherSpringer Science and Business Media Deutschland GmbH
Pages398-414
Number of pages17
ISBN (Print)9783031993527
DOIs
Publication statusPublished - 2026
Event11th International Conference on Information Management, ICIM 2025 - London, United Kingdom
Duration: 28 Mar 202530 Mar 2025

Publication series

NameCommunications in Computer and Information Science
Volume2540 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference11th International Conference on Information Management, ICIM 2025
Country/TerritoryUnited Kingdom
CityLondon
Period28/03/2530/03/25

Keywords

  • AutoML
  • Consumer Well-being
  • ICE
  • Marketing
  • SHAP

Fingerprint

Dive into the research topics of 'Exploring the Determinants of Well-Being: Insights from SHAP and ICE Analyses'. Together they form a unique fingerprint.

Cite this