Please use this identifier to cite or link to this item:
http://acervodigital.unesp.br/handle/10400.2/1282
- Title:
- Análise de agrupamento incremental : segmentação de pontos de retalho
- 1646-6330
- O presente artigo apresenta um estudo sobre a utilização do algoritmo incremental Two-Step para identificar grupos homogéneos de pontos de venda de um universo de pontos de retalho para produtos alimentares congelados. Com este trabalho pretende-se efectuar a segmentação desse universo, de forma a suportar a tomada de decisão por parte dos gestores de marketing e vendas. Os grupos são identificados utilizando informação proveniente de um data warehouse que agrega dados sobre as características de cada ponto de retalho e as vendas que origina. Os resultados obtidos permitiram a identificação de 4 grupos, cujo perfil foi traçado e avaliado mediante o recurso a alguns testes de hipóteses.
- Abstract: The present study concerns the utilization of the Two-Step incremental clustering procedure to identify homogenous clusters of retail points that support the distribution network of frozen food products. The work is aimed to segment the retail points’ iverse, in order to support the marketing and sales’ decision making. The segmentation is based on information stored in a data warehouse that includes stores characteristics and sales performances of each retail point. The results obtained allowed the identification of 4 clusters which profile was identified and evaluated using hypothesis tests.
- 2007
- Jesus, Neuza Brito de; Cardoso, Margarida G. M. S. - Análise de agrupamento incremental : segmentação de pontos de retalho. "Revista de Ciências da Computação" [Em linha]. ISSN 1646-6330. Ano 2, vol. 2, nº2 (2007), p. 26-38
- Universidade Aberta
- Data mining
- Agrupamento
- Segmentação
- Retalho
- Clustering
- Segmentation
- Retail
- REA
- openAccess
- outro
- http://repositorioaberto.uab.pt/handle/10400.2/1282
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.