recurrence-mimicking-learning • Science 44%

Hyper-efficient Offline Recurrent Reinforcement Learning Algorithm. It solves decision path of any length without sequential processing. Implemented for Sharpe Ratio optimization as a base problem.