Avaliação de Modelos Preditivos em Estatísticas Esportivas Ao Vivo: Uma Abordagem de Dados em Tempo Real

Abstract

Sports prediction has expanded remarkably with the advancement of technology and artificial intelligence, which encourages the search for more refined methods of data collection and pattern recognition. This study focuses on the analysis of models and techniques such as Neural Networks and AutoML to understand the occurrence of events in live matches, adopting a large-scale data approach. It was observed that feature engineering and defining time windows before events are crucial for model performance.