What is it?
This research explored the possibilities of ‘machine learning’ in predicting the inflow into or outflow from the social domain in the municipality of Eindhoven. For this, a data set was used which combines the variables from a file compiled for the Association of Dutch municipalities (VNG Realisatie) with inflow data over 2015. The analysis has produced a number of profiles of social assistance entrants. Subsequently, the prevalence of these profiles in the municipality was determined.
The research comprises two components:
- First, risk profiles were developed for the social assistance inflow throughout the Netherlands. Groups at higher or lower risk of entering social assistance were made on the basis of person variables and life event variables. This results in profiles of persons at a high inflowing risk in the social domain.
- Then, these profiles are used to map the number of residents per profile in the municipality, and to determine the residents’ inflowing likelihood per profile. In addition, it indicates the number of entrants per profile in the municipality.
This study is possible for each scheme known in the social domain; schemes regarding participation, youth and the Social Support Act (Wmo).
What are the benefits?
Insight into which groups of persons in the municipality are most likely to enter the social domain, how many residents are involved and where they are in a municipality.