Análisis de canasta de mercado en supermercados mediante mapas auto-organizados
Fecha
2021-11-08
Profe guía
Perfil ORCID
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidade Federal do Paraná
ISBN
ISSN
2237-826X
ISSNe
Resumen
Introducción: Una cadena importante de supermercados de la zona poniente de la capital de Chile, necesita obtener información clave para tomar decisiones. Esta información se encuentra disponible en las bases de datos, pero necesita ser procesada debido a la complejidad y cantidad de información, lo que genera una dificultad a la hora de visualizar. Método: Para este propósito, se ha desarrollado un algoritmo que utiliza redes neuronales artificiales, aplicando el método SOM de Kohonen. Para llevarlo a cabo, se han debido seguir ciertos procedimientos claves, como preparar la información, para luego utilizar solo los datos relevantes a las canastas de compra de la investigación. Luego de efectuado el filtrado, se tiene que preparar el ambiente de programación en Python para adaptarlo a los datos de la muestra, y luego proceder a entrenar el SOM con sus parámetros fijados luego de resultados de pruebas. Resultado: El resultado del SOM obtiene la relación entre los productos que más se compraron, posicionándolos topológicamente cerca, para conformar promociones y bundles, para que el retail mánager tome en consideración. Conclusión: En base a esto, se han hecho recomendaciones sobre canastas de compra frecuentes a la cadena de supermercados que ha proporcionado los datos utilizados en la investigación.
Market baskets, or shopping baskets, are all products that are purchased by a customer at a certain point in time. The analysis of these baskets allows us to know the preferences of our customers, which can be used for various operational, advertising, strategic and logistical purposes. Best of all: it allows us to "predict" their future preferences. We present the case study of an important supermarket chain in the western part of the capital of Chile that needs to obtain key information about their customer market baskets to make decisions. Method: Data preprocessing was performed in order to transform the original data into market baskets. We clustered market baskets using artificial neural networks of the self-organizing maps (SOM) class. The use of the algorithm included the search for the best hyperparameters: grid size and learning rate. Result: The result of the best SOM found identifies six clusters of market baskets, each based in one predominant product, and identifies the products most related to them. Conclusion: Recommendations on frequent shopping baskets have been made to the supermarket chain that has provided the data used in the research
Market baskets, or shopping baskets, are all products that are purchased by a customer at a certain point in time. The analysis of these baskets allows us to know the preferences of our customers, which can be used for various operational, advertising, strategic and logistical purposes. Best of all: it allows us to "predict" their future preferences. We present the case study of an important supermarket chain in the western part of the capital of Chile that needs to obtain key information about their customer market baskets to make decisions. Method: Data preprocessing was performed in order to transform the original data into market baskets. We clustered market baskets using artificial neural networks of the self-organizing maps (SOM) class. The use of the algorithm included the search for the best hyperparameters: grid size and learning rate. Result: The result of the best SOM found identifies six clusters of market baskets, each based in one predominant product, and identifies the products most related to them. Conclusion: Recommendations on frequent shopping baskets have been made to the supermarket chain that has provided the data used in the research
Descripción
Lugar de Publicación
Sponsorship
Citación
AtoZ: novas práticas em informação e conhecimento, Vol. 10, N° 3, (2021).
Palabras clave
Licencia
Atribución-NoComercial-CompartirIgual 3.0 Chile (CC BY-NC-SA 3.0 CL)