Overcoming Data Loss in Wearable Disease Detection with GAN-Based Imputation ↗
Studies generative reconstruction of missing wearable heart-rate data and its potential to support disease-signal detection.
Methods, findings & limitations ↗Data & analysis
Assess incomplete data and investigate patterns in connected health and environmental observations.
Published imputation research; broader AI approaches remain research directions
Measurements & outputs
Research workflow
Audit data quality and define the analysis question before model development.
Evaluate imputation and prediction against suitable reference data, with separate validation and uncertainty assessment.
Interpretation & current limits
Documented applications & research context
Links identify relevant studies and settings. Configuration and implementation stage differ between projects.
Publications & evidence
Read the study-specific methods, results and limitations. The framework paper describes the wider approach.
Studies generative reconstruction of missing wearable heart-rate data and its potential to support disease-signal detection.
Methods, findings & limitations ↗Shows how Nouna population, weather, wearable and indoor climate datasets can be linked in a knowledge graph.
Methods, findings & limitations ↗Methods in exchange
Source & scope
Content and images draw on Sandra Barteit’s Freiburg lecture, The Impact of Climate Change on Human Health: Lessons from Low- and Middle-Income Countries, slides 40, 51, 56, 57, 58, 62. Publication pages provide the corresponding research evidence. Reviewed 12 November 2026.
Images extracted from the lecture retain the original device or research illustration. Product images identify examples used in the source material.