We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Automatic parameter tuning for functional regionalization methods.
- Authors
Casado ‐ Díaz, José Manuel; Martínez ‐ Bernabéu, Lucas; Flórez ‐ Revuelta, Francisco
- Abstract
The methods used to define functional regions for public statistics and policy purposes need to establish several parameter values. This is typically achieved using expert knowledge based on qualitative judgements and lengthy consultations with local stakeholders. We propose to support this process by using an optimization algorithm to calibrate any regionalization method by identifying the parameter values that produce the best regionalization for a given quantitative indicator. The approach is exemplified by using a grid search and a genetic algorithm to configure the official methods employed in the UK and Sweden for the definition of their respective official concepts of local labour markets.
- Subjects
UNITED Kingdom; SWEDEN; LABOR market; MATHEMATICAL optimization; GENETIC algorithms; ALGORITHMS
- Publication
Papers in Regional Science, 2017, Vol 96, Issue 4, p859
- ISSN
1056-8190
- Publication type
Article
- DOI
10.1111/pirs.12199