Financial portfolio management is the process of constant redistribution of a into different financial products. This paper presents a-model-free Reinforcement Learning framework to provide a deep machine solution to the portfolio management problem. The framework consists the Ensemble of Identical Independent Evaluators (EIIE) topology, a-Vector Memory (PVM), an Online Stochastic Batch Learning (OSBL), and a fully exploiting and explicit reward function. This framework is in three instants in this work with a Convolutional Neural Network(CNN), a basic Recurrent Neural Network (RNN), and a Long Short-Term Memory(LSTM). They are, along with a number of recently reviewed or published-selection strategies, examined in three back-test experiments with a period of 30 minutes in a cryptocurrency market. Cryptocurrencies are and decentralized alternatives to government-issued money, with as the best-known example of a cryptocurrency. All three instances of framework monopolize the top three positions in all experiments, other compared trading algorithms. Although with a high rate of 0.25% in the backtests, the framework is able to achieve at 4-fold returns in 50 days.
No takes yet. Share an insight, caveat, or question.
Jiang et al. (2017) studied this question.