Big Data & Machine Learning

Game DNA: Predicting football results

ImpactHub - Atlantic room

18th November, 15:00-15:30

This talk represents a deep dive into an approach for predicting the results of football matches in the English Premier League (EPL) using Machine Learning algorithms. From historical data, a feature set was created that includes game data and historical team performance. This is considered as a virtual DNA of football games and teams, where features are regarded as genes that behave uniquely from one match to another. Predictions include the following: winner of the game (or draw), number of yellow cards, fouls committed, shots on target, number of corners, and number of goals (both teams combined)

Andrei Bâcu

Paddy Power Betfair

Senior Data Developer with more than 5 years of experience in Business Intelligence & Data Warehousing, working with Microsoft, Oracle, and Amazon Web Services technologies. Focused on solving complex data-driven business cases with innovative technologies that fall under the umbrella of Big Data, BI & Data Warehousing, and Data Science. Equally passionate about Kaggle competitions, Data Hackathons, and Machine Learning challenges.