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: Optim who wrote (283)	10/22/1998 10:35:00 AM
From: CatLady	Respond to of 805

Optim, "I apologize if I lead anyone offtrack with my previous recommendation and I will be sure to scrutinze my results more thoroughly in the future." No harm done, that was, in fact, a useful little experiment to go through. Like you, I've noticed that nets train faster with fewer neurons. I've been training starting with 6 neurons and using the add neurons as needed setting. All I can say is that some nets eventually add all the way up to my software limit of 40, others don't. Using settings geared towards fast training lets me experiment with different inputs and influence settings and get feedback within minutes. But once I've settled on good inputs, accuracy rather than speed becomes the dominant concern. CL

To: Optim who wrote (283)	10/22/1998 10:41:00 AM
From: Jay Hartzok	Respond to of 805

Optim, I have trained a number of nets using less than 20 and find that within 20 minutes I know if a net is good or not. I always do that. If the net looks promising I halt the training process, adjust the neurons and start training again. The net sort of "blinks" and seems to lose it's way, when you hit it with the extra neurons, but it quickly corrects itself, and in a short amount of time you have the stats and chart pattern back again. Jay

To: Optim who wrote (283)	10/23/1998 3:16:00 PM
From: Optim	Read Replies (2) \| Respond to of 805

All, Just a few questions for the thread: 1. When you pick a stock to analyze in NS what do you look for? I mean I look for nice volatile issues that look to have a repeating pattern over time. Resource stocks would be an example. 2. How many people have tried very long training periods? I know the free downloading service only gives 3 years of data, but I have tried to export my metastock (Reuters) data to ASCII files for use in NS. I am testing it currently, with four nets training at home as I write this. 3. Has anyone tried relateds such as the dollar, advance/decline, sentiment, call/put, VIX, etc? I am hoping to play with this a little more this weekend, but I just wanted to see if anyone had tried it. 4. What other methods do people use to confirm buys and sells with NS? I like basic TA, looking for support and resistance levels, trendlines, MACD, etc. 5. Having placed my first two trades based on a NS prediction today, what sort of results have people been able to manage using NS? I am just curious as to what % return one should be able to achieve? Thanks to any that respond. Optim