November 1999 Newsletter

I. Adaptive Net Indicators

Adaptive Net Indicators. The new Adaptive Net Indicators, our new “addon” product for the NeuroShell Trader and Trader Professional, are specially formulated versions of the “genetic” method in the NeuroShell Predictor and Classifier. However, they work just like any other indicator works in the NeuroShell Trader!

Adaptive Net Indicators have an unprecedented set of features that, if you think about it, neither we nor anyone else has ever put in a net, as far as we know:

1. You can set the contribution factors yourself, to tell the net in advance how strongly you want the inputs to be used.
2. With the Trader Pro, you can not only optimize contribution factors, but you can optimize the training set size and how far ahead to predict! Think about that for a minute to appreciate the innovation.
3. These nets retrain on every new bar of data (almost instantly). Therefore, your net is always trained on absolutely the latest data.
4. You can predict or classify.
5. If you classify, you can get “confidence” factors.
6. You can build adaptive moving averages to your own specifications using adaptive net indicators.
7. You can nest these nets (ensemble estimators) even easier than nesting Turboprop 2 nets. That is because you never have to worry about in-sample vs out-of-sample; the adaptive net indicator never uses inputs on which it has already trained. The training set is always the N bars previous to the bar being input. You may also have to think about that for awhile to really appreciate it!

By the way, Adaptive Net Indicators are fast enough to work well in the NeuroShell Daytrader Professional, too.

For more detail on Adaptive Net Indicators, visit www.neuroshell.com and click the “Addons” sidebar.

II. Special Referral Bonus to NeuroShell Trader and Trader Pro Users

Special Referral Bonus to NeuroShell Trader and Trader Pro users. During the months of November and December 1999, we will give a free copy of Adaptive Net Indicators to any user that refers someone to us who purchases the NeuroShell Trader, Trader Professional, or Daytrader Professional (assuming the latter is out before the end of the year!) The new purchaser must state WHEN PURCHASING that he/she was recommended by you.

III. More on Addons

More on Addons. While we are on the subject of addons, we urge all NeuroShell Trader and Trader Professional owners to read about the other addon products for the NeuroShell Trader and Trader Pro. Our users have been saying good things about the Mesa and Wavefin products.

IV. Chicken Feed

Chicken Feed. Yes, literally. Dr. William Roush and his colleague Terri Cravener at the Penn State Department of Poultry Science are pioneers in applying neural networks to analyze poultry feed and other aspects of the health of “broiler” chickens. Their extensive research has placed them in the forefront of the scientific application of non-linear modeling in their field. Dr. Roush also speaks about and teaches neural network applications. Here are the titles of just a few of their papers:

Artificial Neural Network Prediction of Amino Acid Levels in Feed Ingredients.

Probabalistic Neural Network Prediction of Ascites in Broilers Based on Minimally Invasive Psychological Factors.

Evaluation of Broiler Growth Velocity and Acceleration in Relation to Pulmonary Hypertension Syndrome (PHS).

V. NeuroShell Trader release 2.3

NeuroShell Trader release 2.3. Release 2.3 is almost ready for our users. We hope to have it available for free download within a few days, and we apologize for the delays getting it out, because we know quite a few users have problems with 2.1 that 2.3 will fix, especially External DLL Call and TradeStation 2000 problems. However, our beta testers of release 2.3 kept finding bugs, and we wanted to get them ALL out. Watch http://www.ward.net for its availability.

VI. Sleep Apnea

Sleep Apnea. Dr. Wayne Danter and his colleagues at the University of Western Ontario used the NeuroShell Classifier to develop a predictive model for diagnosis of obstructive sleep apnea (OSA). Their GRNN model was able to accurately rule in OSA from clinical data, and GRNN did not misclassify patients with moderate to severe OSA. In their study, they further found that use of the neural network could have reduced the number of polysomnography (PSG) studies. If you want to read about Dr. Danter’s work, here’s where to look:

“Neural Network Prediction of Obstructive Sleep Apnea From Clinical Criteria.” CHEST August 1999; 116:409-415.

VII. Thanksgiving Holiday

Thanksgiving Holiday. Ward Systems Group will be closed November 25 and 26 to allow our employees to celebrate the Thanksgiving holiday.

Was this article helpful?

Related Articles