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:

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.

However, the participants had the opportunity to submit the same code (provided it is correct) to the Q21 (in other words no correlations check are performed).

It is very cumbersome to remove strategies from the web page, so they are currently displayed.

If your strategy will continue to perform good and we believe it can be traded, we will contact you.

Other strategies have been funded, and agreements are in place between quantiacs, investors and quants. We cannot disclose more details now, sorry.

5M USD is a reasonable capacity a strategy could handle, yes.

Incorporating seed initialization into your PyTorch code ensures reproducibility by making the random number generation predictable. This involves setting seeds for the PyTorch engine, NumPy, and the Python random module if you're using it. Below, I'll show you how to integrate seed initialization into your existing code. Remember, while this can make your experiments more reproducible, it does not guarantee identical results across different hardware or PyTorch versions due to the inherent nondeterminism in some GPU operations.

I think I won't be available next week. If you have any more questions, don’t expect an answer from me next week.

it simply means that your system should make a new global high before you are entitled for a payment. If your system makes 1000 at the end of January, 800 at the end of Februray, 900 at the end of March and 1100 at the end of April, your profit will be generated at the end of April and they will amount to 1100-100=100;

no, once the system starts being traded, it will be traded in the form it was at submission time, i.e. the quant will not be allowed to update parameters/change details. Of course a submitted system can have an adaptive logic, by changing parameters according to the value of some meta-indicator. If the quant believes there is some big change to be made, it is ok to re-submit the changed system, but it will need again to accumulate a track record before being traded.

https://quantiacs.com/documentation/en/data/fundamental.html

best regards

However, it broke my old strategies. So he further suggested to:
https://quantiacs.com/community/topic/46/macroeconomic-data-with-quantiacs/5?_=1619556376479
And my old strategies are running again.

Ok, so looks like for now for all strategies without a state I have to output and pass None to weights for the state & pass a state variable to the strategy.
Will keep you updated. Thanks for looking into the issue.

Example with two assets

Console output

Example 1: The absolute exposure is 2 > 1, so every weight is divided by 2, yielding 0.75 and 0.25.

Example 2: The exposure is 0.5 < 1, so the scaling factor is set to 1 and the weights stay 0.3 and –0.2.

In short, even if your strategy outputs more than 100 % exposure, normalize always scales it back so the total absolute exposure never exceeds 1—preventing leverage on the Quantiacs platform.