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: Bill Scoggin who wrote (200)	10/8/1998 12:13:00 AM
From: Len Giammetta	Read Replies (1) \| Respond to of 805

Bill, as a matter of fact I believe that Neurostock can do that, but you have to purchase the Genetic Algorithm model. If your interested in another program that uses Genetic Algorithms, you can download a shareware program called NGO. Essentially you build the nets from scratch, preprocess the information, set the parameters, and it will kick out the best ten nets it finds, usually in less than a few hours for some simple nets. Both Jay and I experimented for a few months with it. The kicker is that the best nets ( the fittest) are not necessarily the most accurate, and no matter what I tried I was never satisfied with the results. Jay even experimented with inputting the those weighted inputs you mentioned into his Neurostock models, but I don't think the experiment was successful. He can tell you more about that than I can. Having given NGO more than a fair chance, I really appreciate the relative simplicity, and accuracy of Neurostock. According to an article I read, while memorizing the data is a consideration, most current programs have safeguards to minimize that possibility.

To: Bill Scoggin who wrote (200)	10/8/1998 1:22:00 AM
From: Jay Hartzok	Read Replies (1) \| Respond to of 805

Bill, I am not throwing in everything, including the kitchen sink, to formulate the net. The old nets I made, that seem to take forever to train, have Neuro use all of the available data instead of just selected parts, with the influence settings basically set, with some exceptions, as recommended in the help files. All that I am reporting are observations, that seem to show that the old nets with the extra data available for processing. are handling the current market situation better than my newer models that have only certain selected data, with all influence settings that are "by the book", available for Neuro to process. Jay