Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : Neural Nets - A tool for the 90's -- Ignore unavailable to you. Want to Upgrade?

To: Optim who wrote (584)	5/6/1999 11:59:00 AM
From: Brian Hornby	Read Replies (1) \| Respond to of 871

Thanks. I use only 13 models for the system. But may add more. Sometimes the averaging makes the system model actually perform worse then the individual ones, which is why I think that the Meta model approach may be better. However I have not seen much user feedback on the meta models. Certainly my meta models are better than the original. Also they are consistent, all meta models generated were profitable. As they were generated just recently though there is not much history, versus the standard models generated on Feb 20. As you know, out of sample performance is one standard, however that doesn't mean a model performing well in out of sample test will continue to do so, people select the best out of sample performance for their models. Subsequent real time trading is the only true out of sample test.