Servicios Personalizados
Articulo
Indicadores
Links relacionados
- Citado por SciELO
- Similares en SciELO
Bookmark
Revista de Investigación Estudiantil Iluminate
versión impresa ISSN 2415-2323
Resumen
CAUNA COARITE, Rodrigo Alejandro y CRUZ BENITEZ, Christian Gerardo. Classification of clients with neural networks to prevent dropouts to a service. Rev. Inv. Est. I. [online]. 2019, vol.11, n.1, pp. 33-44. ISSN 2415-2323.
Abstract The systems, formed by several sets of elements, tend to lose stability when elements are separated from the system, to solve a neural network will obtaín separation patterns with a datábase of afñliates who are about to change medical insurance and classify them to anticípate their separatíon with the insurance, Using quantitative methodologies, the network will classify afñliates based on whether they will abandon or continué to use insurance, After testing, the network worked with a hit rate of 91,43%, with another similar datábase the hit rate was 90,18%, With these results it was concluded thatthe networkworks effectively forany otherservice aslongasithas information about this type of users,
Palabras llave : Neural networks; separations; systems.