CNN-DRL with Shuffled Features in Finance

Montazeri, Sina, Mirzaeinia, Akram, Mirzaeinia, Amir

arXiv.org Artificial Intelligence 

FinRL-Meta offers a standardized benchmarking system, enabling the evaluation of different reinforcement Deep Reinforcement Learning (DRL) has emerged as a learning algorithms on common financial tasks. The authors powerful paradigm in the realm of finance, revolutionizing also introduce several benchmark datasets and demonstrate the way investment strategies are formulated and executed. It the effectiveness of FinRL-Meta by training agents on these combines the strengths of deep learning, a branch of artificial datasets. The platform serves as a resource for the development intelligence, with reinforcement learning, a framework for and comparison of data-driven reinforcement learning learning optimal decision-making policies through trial and techniques in the context of financial markets.