NeuroStock Message Board

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

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next

To: Len Giammetta who wrote (455)	12/1/1998 7:33:00 PM
From: Jay Hartzok	Read Replies (1) of 805

Len, Does this mean that the influence period does,'t influence the training cycle, just the predictions? Are you confusing "predictions" as pertaining only to untrained data, the verify period and beyond? During the training process Neuro makes a prediction for every day of trained data. If the prediction is accurate this reinforces the current model. If it is an error then Neuro retrains the data and looks for the data relationship that will make a correction. Since there is a prediction for every day of the training period and verify period, the influence settings tell Neuro how far back to look in the data to formulate the prediction for each day. The only difference I can see between the verify period and the training period is that the predictions in the training period supposedly have been corrected if they were at one time wrong. Those in the verify period don't get corrected. Those predictions are based on the model that the training period created. If that's the case why not set all the influences for 30 days? Is the answer to this one, "training the net would take too long"? Jay

Report TOU Violation

Share This Post

Public Reply Prvt Reply Mark as Last Read File

Previous 10 Next 10 Previous Next