Skip to content

SIPHER7 Intervention

Joao A. Duro requested to merge JoaoEI into main

Added a new SIPHER7 intervention model (namely sipher7_intervention) to the interventions module.

  • Interventions are made to six components of the equivalent income metric, and these are: (1) physical health, (2) mental health, (3) loneliness, (4) employment, (5) housing, and (6) safety. Currently intervention with respect to the income is not considered in this model.
  • Only one intervention can be selected for each location by using the integers shown above, while 0 is reserved for no intervention.
  • The intervention model determines the following performance criteria: (f1) the number of locations with intervention; (f2) the mean of the equivalent income in the whole jurisdiction (e.g. LAD of Sheffield); (f3) and an inequality metric (in this case it is the difference between the maximum and minimum equivalent income (averaged), determined with respect to each location).
  • Each individual in the population has an equal chance of benefiting from an intervention, and this chance is equal to the probability (pmax)---a parameter of the model. By default it is set to 10%.
  • Due to the use of probabilities, there is script that quantifies the robust performance of each solution. This script calls the sipher7_intervention function 10 times and takes the 90% percentile of the performance with respect to f2 and f3.

Merge request reports