Institute of Housing Research · Landshut
Deutsch
Institute of
Housing ResearchIfWf

Institute of Housing Research

Modelling care and transitions

Computational models help make assumptions about support needs, participation and possible actions explicit and testable.

Starting with the care relationship

A transition from hospital to home involves at least the person receiving support and another person involved in care. Abilities, available help and readiness may develop differently. We therefore consider multiple perspectives and their trajectories together.

States are not directly observable

Burden, confidence or readiness cannot be read directly from individual clicks or use events. Probabilistic state models can link observations to explicit uncertainty. Our development work focuses on scenarios with incomplete information and repeated decisions, including partially observable Markov decision processes (POMDPs).

A scenario to investigate

  1. Before discharge: describe activities, support needs and both participants’ perspectives.
  2. During transition: compare support options such as training, technical adaptations or additional assistance.
  3. At home: use feedback and observable events to update assumptions.
  4. Evaluate: does a recommendation remain stable when assumptions change, and does it fit participants’ assessments?

Related texts