Machine learning hedge strategy with deep Gaussian process regression

An optimal hedging strategy for options in discrete time using a reinforcement learning technique

CLICK HERE TO DOWNLOAD THE PDF

When applying machine learning to trading strategy, two inevitable practical issues are achieving interpretable results and securing robustness to market changes. To overcome these challenges, Yoshihiro Tawada and Toru Sugimura propose a new method to obtain a hedge strategy for options by applying Gaussian process regression to the policy function in reinforcement learning. The method is tested using typical option schemes with a transaction cost and compared