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 : NeuroStock -- Ignore unavailable to you. Want to Upgrade?

To: Len Giammetta who wrote (617)	1/14/1999 3:31:00 PM
From: Jay Hartzok	Read Replies (1) \| Respond to of 805

Len, The net won't actually see the errors in the trained data until those days of missing data become part of the net. But we're not really interested in what the net sees in the training period, we're interested in the data in the verify period which is where the missing data is. If your net is setup to look back and analyze the data for the last 15, 20, or 30 days in order to make the next prediction, then you can't believe any current prediction because all of the data is not there. So it seems to me that, depending on the highest number of influence days that is contained in the net for any one input, we have to wait until all of those days expire before we will again get any true predictions. For most of my nets this will occur during the last week of this month and for some of my nets won't occur until the second week of next month. Bill may be right when he said Neuro replaces missing data with the previous days data, but is bad data what we want to place our trust in? Jay