A carregar...
Publicação
Competing forecasting models to study crisis periods : the case of sweet snacks sales
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.05 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
The COVID-19 pandemic significantly affected the purchasing behaviour of
Portuguese families, compelling them to reduce their shopping expenditures. This
socioeconomic crisis necessitated that food retailers adapt their strategies to evolving
consumer preferences, emphasizing digitalization, sustainability, and safety. This study
examines the sales evolution of the Sweet Snacks category at two major Portuguese retail
banners1
from January 2018 to June 2023, segmented into three forecasting periods: precrisis, crisis and post-crisis. The project’s primary objective is to infer forecasting models
for these periods, using the ARIMA and Prophet time series models, and compare them
to assess consumer preference changes. Additionally, this work forecasts Sweet Snacks
sales beyond June 2023 to extend the appraisal of sales performance in the post-crisis and
detect potential anomalies in the sector. Using the CRISP-DM methodology, the research
developed an integrated BI solution, employing Power BI for data preparation and R
Studio for multidimensional data modelling and forecasting analysis. In the pre-crisis
period, Sweet Snacks sales progressively increased until the onset of lockdown, declining
in the crisis period. At the end of the crisis, consumption patterns normalised, but postcrisis, retailers diverged due to their adaptability to new trends. The results indicate that
while ARIMA models generally offer higher accuracy, Prophet provides more precise
future forecasts. ARIMA predicts a steady future trend, whereas Prophet captures postcrisis sales patterns more effectively. This project’s main contribution is the development
of a BI solution and a comprehensive forecasting report for a Consulting organisation in
the Food Retail sector.
Descrição
Mestrado Bolonha em Data Analytics for Business
Palavras-chave
Time Series Forecasting Crisis periods Sweet Snacks sales ARIMA Prophet CRISP-DM
Contexto Educativo
Citação
Barradas, Sara Cristina Coelho (2024). “Competing forecasting models to study crisis periods : the case of sweet snacks sales”. Dissertação de Mestrado. Universidade de Lisboa. Instituto Superior de Economia e Gestão
Editora
Instituto Superior de Economia e Gestão