Berlin 2024 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

AKPIK: Arbeitskreis Physik, moderne Informationstechnologie und Künstliche Intelligenz

AKPIK 3: Poster

AKPIK 3.6: Poster

Thursday, March 21, 2024, 11:00–14:30, Poster B

Unraveling Chronic Disease Relationships: A Comparative Analysis of Clustering Algorithms on the DHS 2019-2021 Indian Dataset — •Jannis Demel, Anna Nitschke, Carlos Brandl, Jonathan Berthold, and Matthias Weidemüller — Physikalisches Institut, Ruprecht-Karls Universität Heidelberg, Im Neuenheimer Feld 226, 69120 Heidelberg, Germany

Clustering algorithms play a pivotal role in unsupervised machine learning, offering a systematic approach to discerning patterns and associations within intricate datasets. The focus of application is the DHS 2019-2021 dataset from India, utilizing biomarkers to identify clusters of individuals. We compare four distinct clustering algorithms (K-means, DBSCAN, GMM, and HDBSCAN) and evaluate their performances from a data analytics and medical point of view. This approach aims to unveil novel insights into the relationships between chronic diseases. Through this poster, we contribute to a nuanced understanding of chronic diseases in India, offering valuable insights into the practicality of clustering algorithms in healthcare analytics.

Keywords: Clustering Algorithms; Unsupervised Machine Learning; Global Health