A carregar...
Publicação
A redefinition of iFood’s Coverage Health - Score including qualitative features
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.32 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Over the last years iFood’s coverage strategy focused on the expansion of their coverage
to new areas and increasing the quantity of restaurants in existing areas. Both of these
strategies have been effective as they resulted in a higher performance in orders per
inhabitant for the areas that the strategy was applied to. When areas became covered
well, the rate of the performance increase tended to slow down and a saturation point
was about getting reached. Therefore, iFood needed a new strategy to keep the performance
of well covered areas growing after saturation. The focus of this internship
report was to find the new coverage strategy for iFood. As the quantity of restaurants
was no longer the focus point, it was relevant to look into quality attributes that make
up an area. An exploratory data analysis of coverage attributes was conducted to find
these quality attributes. By doing this it was found that there were four attributes
that show potential in increasing the performance of well covered areas. Each of these
attributes were analyzed to come up with thresholds that would decide whether an
area is well covered in regard to this attribute. The attributes and their individual
thresholds were then used to develop a new model which uses the QS. The QS consists
of the four attributes and their thresholds as well as an assigned weight to mark the
importance of the individual attributes. Lastly, the new quality strategy was incorporated
into the existing strategies for iFood to come up with a new version of the
CHS.
Descrição
Internship Report presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics
Palavras-chave
Growth Strategy; Supervised learning; Unsupervised Learning; Feature Analysis;