PGPortfolio: Policy Gradient Portfolio, the source code of "A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem"(https://arxiv.org/pdf/1706.10059.pdf).
Hi! Thanks for your great paper and code.
In the backtest part, the agent uses 'decide_by_history' to make decisions. I was wondering whether the agent is able to make decisions and learn at the same time, by changing the 'tflearn.is_training' in the 'decide_by_history' to True?
Hi! Thanks for your great paper and code. In the backtest part, the agent uses 'decide_by_history' to make decisions. I was wondering whether the agent is able to make decisions and learn at the same time, by changing the 'tflearn.is_training' in the 'decide_by_history' to True?