A carregar...
Publicação
Machine Learning: Applying Genetic Algorithms for Budget Management & Team Composition in Fantasy Premier League (FPL)
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 1.66 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
This thesis explores how genetic algorithms (GAs) can be used to optimize team selection in
the Fantasy Premier League (FPL), an online multiplayer fantasy game. The main goal is to find
the best possible teams that maximize overall performance, considering constraints like
budget limits and player positions. The research looks at various configurations of GA
operators, using different selection strategies, crossover methods, and mutation techniques.
Specifically, it examines rank selection, tournament selection, and roulette wheel selection for
selection strategies; single-point, double-point, and uniform crossover methods; and scramble
mutation, swap mutation, and inversion mutation for mutation techniques. A total of 10
different GA configurations are tested to see which one works best for creating teams in FPL.
The results show that the best configuration involves tournament selection with uniform
crossover and scramble mutation. This setup consistently achieved the highest fitness scores,
indicating it performs best in team optimization. The genetic algorithm was tested using
historical data from the 2023-2024 FPL season. The fitness function evaluated team
configurations based on player performance (points per game), budget constraints, and other
game rules like player positions.
The findings demonstrate that genetic algorithms are effective for the complex task of FPL
team selection. This thesis adds to the field of Machine Learning in Fantasy Sports by
comparing different GA setups and pointing out the most effective strategies for improving
team performance. Future research could look into adding more constraints and performance
metrics to further enhance the optimization process.
Descrição
Dissertation presented as the partial requirement for obtaining a Master's degree in Data Science and Advanced Analytics, specialization in Data Science
Palavras-chave
Genetic Algorithms (GAs) Fantasy Premier League (FPL) Machine Learning (ML) Optimization Budget Team Selection Football SDG 4 - Quality education SDG 9 - Industry, innovation and infrastructure