| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 941.83 KB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Football players’ market values are intensively researched as the focus gradually shifts from
qualitative approaches and subjective perception to data-driven estimation. Previous work drew
on player characteristics and top-level performance metrics but only few included injuries. This
thesis resorts to data from Opta and Transfermarkt and leverages MLR and machine learning
approaches to derive market values.
The LightGBM and MLR models predict market values accurately and identify direct impacts
of injuries. Additionally, I found that player positions matter for injury incidence and severity.
Descrição
Palavras-chave
Football Market values Injuries Machine learning LightGBM MLR Opta Transfermarkt
