Public case study

Health DKM

Health DKM is an applied case study in turning longitudinal Health and Retirement Study data into decision-ready system intelligence. The aim is to provide a stronger analytical structure for understanding change, risk, and intervention timing.

Problem

Longitudinal health datasets are rich, but most teams still interact with them through fragmented analytics and flat reporting.

Approach

Health DKM turns the data into interpretable state models, trajectory structure, and intervention-relevant views.

Value

Decision-makers get a more legible picture of progression, transition pressure, and what kind of action may matter next.

Representative outputs

Snapshots from the toolchain.

Participant-level data view used in the Health DKM toolset for HRS cohort analysis — Participant-level view for cohort filtering and longitudinal feature inspection. Click image to expand

Cognitive decline surface output from Health DKM showing multidimensional risk structure — Surface output highlighting transition pressure, trajectory shape, and intervention timing. Click image to expand

3D manifold studio visualization generated from Health DKM modeling outputs — 3D manifold studio view for geometry review and interpretability analysis. Click image to expand

Workstreams

How the program develops.

Cohort and variable harmonization across waves

Latent state and transition structure estimation

Robustness and alternative-assumption testing

Decision-facing briefs and reproducible artifacts

Request project brief

Discuss collaboration