Not logged in
Viewing 5 posts - 1 through 5 (of 5 total)

Forums>StrategyQuant>General Discussion>About deep reinforcement learning.

  • #258311 |
    Customer
    251 Posts

    I found an interesting open source project that uses deep reinforcement learning to generate trading strategies. Can SQ learn some useful things from that?

    https://towardsdatascience.com/using-reinforcement-learning-to-trade-bitcoin-for-massive-profit-b69d0e8f583b

    #258318
    Mark Fric
    Administrator
    1172 Posts

    interesting article. We did some experiments with deep learning but didn’t come up with anything robust. This is one of the reasons why Neural Networks are still not added into SQ – every NN model we tried was just curve fitting.

    Mark
    StrategyQuant architect

    #258477
    Customer
    244 Posts

    what about random forest etc?

    #258525
    Customer
    418 Posts

    You have to combine it with rules based entries or it will probably never work and be curvefitting.

    #258526
    Customer
    418 Posts

    So SQx team can add the function to Improver and we can test it out . :)

Viewing 5 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic.