Data Scientist
CurrentDesigned a model to simulate footballer performance as part of a ‘stock’ /football index game. Demand for a stock was simulated using neuro-linguistic programming approach to convert tweets about a a player into positive or negative sentiment scores.Developed a model to predict football scores using an ensemble of Random Forests, General Additive Modelling, and Boosting. The model predicted clean sheets with 80% accuracy and goal scorers with over 88% accuracy.Designed a web crawler to collect footballing statistics from various websites.