AI Factor screens do not seem to be working - no output and error messages

General
1

I tested two completely different trained AI Factors — an existing Predictor that has worked in screens previously, and a newly trained Predictor. For each, I used P123's own Predictor Formula → “run a screen that returns the top decile” link, and other custom screens referencing the predictors.

In both cases, "Run Screen" processes for ~30 seconds and returns no output. Subsequent attempts sometimes report that another Screen request is still in progress.

Conventional (non-AI) ranking systems run normally using the same universe.

Also tried the validation model route - using the exact fx formula, but the exact P123-generated AIFactorValidation() formula returns “Could not find validation model.”

Please help!

Thanks in advance,

Ryan

image
image1017×592 36.1 KB

2

Thanks for reporting this Ryan. We are looking into it.

3

AIFactor predictions have been fixed.

4

AI models are not available in beta version of Asian and Emerging markets , right? It would be useful to quick test and export factors in my case

5

Yes, that is correct - AIFactor, API and Download Factors are not available in the beta environment.

6

Excellent, thanks so much Dan. Confirmed, it works now on my end.

Cheers!

Ryan