Chris Whitrow Email and Phone Number

Applied mathematician, data scientist and former post-doc academic with extensive experience in delivering projects for a wide range of commercial clients. Participant in profitable football betting syndicates. I have been using statistical and ML models to bet successfully on major football leagues since 2006 - in both 1x2 and Asian handicap incl O/U markets. I have also published ground-breaking academic papers in deep learning, machine learning and optimisation. My research has always been guided by theoretical principles taken from information theory, Bayesian inference and empirical psychology. In recent years, I've worked with syndicate partners to trade in Asian handicap football pre-match markets (both results and total goals over/under), making a healthy positive yield, beating the world's sharpest Asian sports books. This requires the creation of predictive models that are among the very best of their kind, which is an ability that comes only with long experience (and learning from mistakes along the way).Recently, I have turned attention to general reinforcement learning algorithms, incorporating temporal-difference learning to improve the Alphazero game-playing algorithm. I am currently training a ResNet to play an ancient Nordic boardgame, using Tensorflow on my local GPU.In addition, I have produced commercial big data models for fraud detection, churn, segmentation / clustering, bond rating, recommendation systems and credit scoring for large banks, retailers and telecom companies, as well as writing peer-reviewed papers on these subjects. My research has included pioneering information-theoretic algorithms for semi-supervised and unsupervised learning. PhD in applications of artificial neural network models.Most of my papers can be downloaded from https://imperial.academia.edu/ChrisWhitrow