QUANTCONNECT COMMUNITY
Hi Everyone,
In this strategy, we apply a Convolutional Neural Network, a Neural Network often used to identify objects from an image, in order to forecast stock prices. I've attached a backtest, but the full writeup for the algorithm can be found here. Furthermore, a video walkthrough can be found here.
Best,
Shile Wen
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.
QuantConnect™ 2024. All Rights Reserved
|
|
Share New Research 
Carpediem911
Nice work Shile! I like the classification approach instead of the time series prediction. To increase the results, did you ever consider to make it more multi-dimensional and include some beside of the bar.data some additional e.g. indicators? And did you ever tested those networks on faster resolutions like resolution.minute? Would be even more interesting… Trying to find some examples, but not successful yet….
Shile Wen
Hi Carpediem911,
That's a good idea, when we extract the timeseries pandas DataFrame, we can apply methods to compute things like SMA's (which can be done with pd.DataFrame.rolling.mean(20) for 20 period SMA, for example). I haven't tested this however.
I tried running it on Resolution.Minute, but the backtest took too long.
Carpediem911
sounds good. Would be great if it could be run on 1 min resolution. Why did it take too long for the backtesting? the completion of training should not be longer if just the resolution is changed?
Carpediem911
And another thing: Did you try your network also on crypto?? Would be even more interesting. However, when trying, we get errors because shorting not possible with cryptos and also by far too high fees, We don't use usually the execution model (but the “old” QCalgorithm) and the “CloseOnCloseExecutionModel” is not explained in the API doc, thus we don't know how to change it to make it working with crypto… THX
Shile Wen
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!