Car category recommendation : a classification approac
Resumo
Resumo: Um sistema de recomendação pode ser implementado em qualquer loja online que ofereça produtos ou serviços com o objetivo de melhor a experiência do usuário e aumentar lucros. O mesmo funciona para lojas de aluguel de carros, nelas um sistema de recomendação também pode ser implementado, de modo que liste o carro que melhor se encaixa com as preferências do usuário. Este projeto tem objetivo de criar um sistema de recomendação de veículos de aluguel usando uma abordagem de classificação com um classificador MLP. Resultados parciais demonstram que esta abordagem tem potencial de sucesso, uma vez que acurácia do classificador em prever qual veículo deve aparecer na primeira posição é de 30% e de 57% de que o veiculo irá aparecer até a tercerira posição Abstract: A recommendation system can be implemented at any online shopping that offers products or services with the objective to offer a better experience to the customer and increase profits. When renting cars, we also can explore the benefits of recommending the best car to match customer wishes. This project works towards building a recommendation system by using a classification framework applying an MLP as a classifier. Partial results have demonstrated that the recommendation can be made with an accuracy of 30% that the best match car will be in the first position of search and that with of accuracy 57% the classifier can predict that the desired car category will be up to the third position.
Collections
- Data Science & Big Data [138]