Samsung field lab voids: comparison of deep learning approaches in demand forecasting for consumer electronic goods

Kuehne, Melina Franziska

http://hdl.handle.net/10362/145523

Utilize este identificador para referenciar este registo.

Nome:	Descrição:	Tamanho:	Formato:
file-1_final_submission_melina_kuehne.pdf		24.6 MB	Adobe PDF	Ver/Abrir

Contacte-nos

Autores

Kuehne, Melina Franziska

Orientador(es)

Han, Qiwei

Resumo(s)

VOIDS is a data science student entrepreneurial project that aims to integrate marketing into demand planning, helping companies to achieve the most accurate way of planning and shaping future demand. The following work applies this vision in a lean agile start-up framework, implementing state-of-the-art deep learning time series forecasting techniques in the business context of the global consumer electronic brand Samsung. The following chapter will present the development of two deep learning models. While the well-known Deep AR model could not improve the overall forecasting accuracy, the newly developed MQCNN showed promising outcomes and resulted in a forecasting accuracy uplift.

Palavras-chave

Marketing Machine learning Business analytics Deep learning Demand planning Consumer electronics Time series forecasting Deepar Mqcnn Demand forecasting