@support Thank you very much once again!

Thanks for the answer!

I still think something is wrong with this correlation checker. I even used this function to randomize the weights a few times, and I got the same correlation error:

def add_random_noise(weights, noise_level=0.01): noise = np.random.uniform(-noise_level, noise_level, size=weights.shape) return weights + noise

I am pretty sure it's impossible to have 90% correlation in this case.

@omohyoid Dear omohyoid,

You don't need to specify the output path nor any other environment variable when submitting your code. If you're using local development option to create a strategy, you should remove setting environment variables manually since those are created automatically after you submit your strategy so no need to worry about that and it is actually advisable to do it that way.

Regards

@support Thanks for ur help

@newbiequant96 Hi, the template is a "working code" still to be finalized and published among the templates in the account area, however the logic behind is strictly multi-pass and a conversion to single pass is not really so straightforward.

@commanderangle Dear commanderangle,

Your strategies are processed in a correct manner, but the reason why you see 0 out-of-sample score is due to the fact that your strategies generate zero weights for all assets for out-of-sample time period. You can check your weights for any strategy by downloading them. There is a download button in the submission logs section.

Regards

Hi @support,
no problem.
I didnt check until now, the accepted strategies do not use machine learning 🙂
I will try out some machine learning strategies in the upcoming days and let you know.
Thanks again and Regards

@buyers_are_back Hello.
Here is a new example of stock prediction using index data.
I recommend using the single-pass version.
https://quantiacs.com/documentation/en/data/indexes.html

@support Hello. My strategy has the id #16934018 and was submitted in early May, but pnl OS has not been updated yet. Please check this issue. Thank you.

@blackpearl Hi, sorry for the missed answer, did you solve the issue? We can see that you managed to successfully submit code to the Q21 contest!

@multi_byte-wildebeest Hi, there are always limitations in a real-world scenario. Here we are running a contest on US stocks only, however imagine to trade assets worldwide. Then you have to take care of different timezones, and once US markets close, you have some hours to generate trading positions for (let us say) Australian markets.

If a ML model takes days to take a decision, that is unusable.

The limitations in place are per point-in-time, it means that for each "pass" the system should take no more than 5 or 10 minutes.

In practice we have also other limitations, because when many systems are in our queue, it can take very long to process all of them.

@illustrious-felice Hi, you can in the notebook, not in the submission area because of resource reasons.

@iron-tentacruel Sorry for the delay in the answer. We recommend conda as we can better track dependencies. With conda you can create locally an environment which mirrors the one on the Quantiacs server and you can work locally as you would on the server. If you need a specific version of a package, please let us know.

@black-magmar You are correct, but this kind of forward-looking is always present when you have all the data at your disposal. The important point is that there is no forward-looking in the live results, and that should not happen as the prediction will be done for a day for which data are not yet available.

@blackpearl Hello. I don’t use machine learning in trading, and I don’t have similar examples. If you know Python and know how to develop such systems, or if you use ChatGPT (or similar tools) for development, you should not have difficulties modifying existing examples. You will need to change the model training and prediction functions.

One of the competitive advantages of the Quantiacs platform is the ability to test machine learning models from a financial performance perspective.

I haven’t encountered similar tools. Typically, models are evaluated using metrics like F1 score and cross-validation (for example, in the classification task of predicting whether the price will rise tomorrow).

However, there are several problems:

It is unclear how much profit this model can generate. In real trading, there will be commissions, slippage, data errors, and the F1 score doesn’t account for these factors. It is possible to inadvertently look into the future. For instance, data preprocessing techniques like standardization can leak future information into the past. If you subtract the mean or maximum value from each point in the time series, the maximum value reached in 2021 would be known in 2015, which is unacceptable.

The Quantiacs platform provides a tool for evaluating models from a financial performance perspective.

However, practice shows that finding a good machine learning model requires significant computational resources and time for training and testing. My results when testing strategies on real data have not been very good.

@commanderangle Dear commanderangle,

If you use ML in your strategy but not select that option we can't guarantee for how your strategy will be evaluated and it could be filtered out.

Regards

@multi_byte-wildebeest Hi. Without an example, it's unclear what the problem might be.

If you use a state and a function that returns the prediction for one day, you will not get correct results with precheck.

This was discussed here: https://quantiacs.com/community/topic/555/access-previous-weights/18

@gjhernandezp Hi,
sorry for delay, you can check this topic:
https://quantiacs.com/community/topic/564/collections-has-no-attribute-iterable

@vg2001 Hi, the limit refers to point-in-time evaluation, namely 10 minutes per point in time, where points in time are the processed historical days.

@vg2001 Sorry for the delay. It is a bit different, we reserve the possibility to scale down volatility to 5%. This can happen if the algorithm (for example) concentrates allocations in a few assets and becomes inherently very risky to be traded.