Outperforming algorithmic trading reinforcement learning systems: A supervised approach to the cryptocurrency market

作者：

Highlights：

• ResNet-LSTM actor as our proposed method for financial trading decision problems.

• Comparison of our method against state-of-the-art reinforcement learning methods.

• Real-world evaluation using cryptocurrency market, surpassing the benchmark.

• Robustness evaluation with and without transaction costs.

• Feature extraction insights using graphical visualization of the layer’s outputs.

摘要

•ResNet-LSTM actor as our proposed method for financial trading decision problems.•Comparison of our method against state-of-the-art reinforcement learning methods.•Real-world evaluation using cryptocurrency market, surpassing the benchmark.•Robustness evaluation with and without transaction costs.•Feature extraction insights using graphical visualization of the layer’s outputs.

论文关键词：Deep neural network,Reinforcement learning,Stock trading,Time series classification,Cryptocurrencies

论文评审过程：Received 31 October 2021, Revised 12 April 2022, Accepted 13 April 2022, Available online 19 April 2022, Version of Record 28 April 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117259