I have created a LSTM deep learning model for trading the forex market(saved in a h5 file).Also the input for the model has to be in a tensor form.I would like to know how to implement this for bactesting on the QuantConnect platform.
I have created a LSTM deep learning model for trading the forex market(saved in a h5 file).Also the input for the model has to be in a tensor form.I would like to know how to implement this for bactesting on the QuantConnect platform.
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.
Hey Sahil,
Is hard to know how to help with the little information you are giving.
But thinking out loud: you can use Lean to port the data in the format you need to feed your model. Then return the output of your model to Lean to execute the orders.
Now I don’t know if the model can be run in the QC platform, it’ll depends on the libraries you need.
The good news is that if you can run a backtest in Lean with the model incorporated, is almost ready to go Live mode.
Well right now i am only concerned about backtests,not live trading.I think i need the keras library.
Also i think instead of backtesting or live trading, i would like to create an indicator using live data stream on QuantConnect.Would that be possible?
I'm not quite sure what do you mean, but in principle, if you can code it, it can be run in Lean.
Please, check the docs for a better understanding of the Lean possibilities.
Well i actually wanna make a signal generating system.It would give me a sort of signal when trade conditions are satisfied.I intend not to have a fully automated trading system but use this tool that i have made as an indicator.I am not able to code a deep learning model on mt4 so i thought it would work here on a python platform.But for that i would need a live data feed.
Sahil,
Yes, you can develop a system like the one you describe in Lean.
Thinking out loud, you have two alternatives:
Use the QC platform by running an algorithm in Live mode. But maybe one of the problems you’ll encounter is that you can send 20 notifications per hour as maximum outside the engine. If that is not a limiting factor for your strategy, I think is the best option.
The another option is run Lean in live mode locally (or in a Virtual Machine). In that case, the data feed will be provided by your broker, you’ll be able to use any library you need and there isn’t a limit to notifications. But requires maintain a small host to run the engine (an old computer with Linux or the smallest AZURE or AWS VM will do the job).
It is totally possible.
At the moment we haven't included keras in out whitelist, but, as you can see in the post JayJayD mentioned above, there is some demand, so we will study this possibility. In order to have a live data feed, you will need to subscribe to a prime account.
After all set, you can receive the signals with our notification system. In live trading mode we allow sending Email, SMS and Web-Hook notifications to notify you of significant events. You can send up to 20 notifications per hour.
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.
To unlock posting to the community forums please complete at least 30% of Boot Camp.
You can
continue your Boot Camp training progress from the terminal. We
hope to see you in the community soon!