In this issue:
I. Opinions of Steve Ward, CEO
II. Special offer for existing users only.
III. Customizing optimization penalties in NeuroShell Classifier
IV. Thanksgiving holiday closures
V. Make sure you continue to get the news
*******************************************************
I. Opinions of Steve Ward, CEO
Item 1. In the December issue of Technical Analysis of Stocks and Commodities the NeuroShell DayTrader Professional is reviewed. In the past we’ve felt that some NeuroShell reviews were just cursory, with little effort by the reviewer to really understand the salient features of our product. However, this time I am pleased to say that writer Dennis Peterson has captured the essence of our software – the things that really make it stand apart. Watch for your copy in mid November, but if you don’t subscribe we’ll probably purchase a link to the review which we’ll place on NeuroShell.com.
Item 2. Speaking of TASC, they’re having their annual software contest again. Back before we went from #2 to #1 in the artificial intelligence category I used to tell people it was just a popularity contest and prospective users shouldn’t place a great deal of emphasis on it for their software buying decisions. Once we became #1 last year, I had an epiphany and realized how valuable the contest is in giving readers the important insight they need to purchase great products! Now I don’t want to change my story again, so I am asking for your help to keep us #1 in artificial intelligence trading software. If you are a TASC subscriber, please go their website now and vote. Here’s how:
Get your copy of Stocks & Commodities magazine and visit www.Traders.com. Look in the upper left hand corner of the web site for the 2004 Readers’ Choice Awards and click on the Vote Now! link.
On the page that is displayed enter your subscriber ID from your magazine address label (look at their example to see which numbers on the label are the subscriber ID). After you are logged in, look at page 3 of the ballot for the Artificial Intelligence Software (Expert, Neural) category. Be sure to mark you ballot for NeuroShell Trader (Ward Systems Group Inc.) Do the same on page 2 of the ballot for the Standalone Analytical Software, $500 and more category. Look for NeuroShell Trader Professional.
Item 3. Many of you have from time to time asked me what type of system I like to trade with. I have usually been pretty open about it, giving you my technique preferences in general. However, a few weeks ago I just decided to write up the whole thing and posted it on www.ward.net with all the appropriate disclaimers. If you feel like you haven’t built anything good yourself, you are welcome to use my ideas.
Item 4. Speaking of ideas, my friend Murray Ruggiero has more than anyone I know, and he has a new book out that many of you may find interesting. It is a plastic ring binder of copies of his articles from Futures Magazine from 1995 to 2003. His articles have covered classic systems, neural nets, genetic algorithms, adaptive systems, game theory, options, and much more. Although many of his techniques will be difficult to implement in NeuroShell unless you can program (there seems to be a lot of TradeStation Easy Language and Excel code), you may still get a great education reading how the techniques work. At very least you’ll get some good ideas that may influence your NeuroShell systems building. Murray has promised to sell the book to NeuroShell users for $20 off the regular $99 price. You can reach him at 800-211-9785. Ask for “Technology in Trading for the New Millennium”.
Thanks for listening.
*******************************************************
II. Special offer for existing users only.
Steve Ward mentioned in his opinion piece the review of NeuroShell DayTrader Professional written by Technical Analysis of Stocks and Commodities magazine. In case that article motivates a REGISTERED EXISTING user of any of our products to upgrade to NeuroShell Trader, Trader Pro, or DayTrader Pro, we’ll make it easier for you. We’ll give you 20% off the purchase or upgrade price until the end of November. However, this is only for people who have already purchased one of our products in the past and who read this newsletter, so please do not tell anyone else about this discount. Here’s the fine print: cannot be combined with any other discount or promotion, and cannot be applied to previous purchases or upgrades. You must specifically ask for this discount when ordering.
*******************************************************
III. Customizing optimization penalties in NeuroShell Classifier
Often when classifying data into categories there are some wrong answers that are worse than others, and some right answers that are better than others. Let’s take an example everyone can understand. Suppose you are a medical researcher building a neural net to screen for a deadly disease. Since the NeuroShell Classifier has a simulated cancer example, and since there are many researchers trying to screen for cancer, we’ll use cancer. You (the researcher) will probably have two categories: Positive and Negative.
If the NeuroShell Classifier gives a high probability that the patient is positive based on the results of tests input to the net, the screening doctor will want to send the patient for a biopsy to confirm. If the probability is high that the patient is negative, the screening doctor will send the patient home with an appointment for another checkup next year.
Suppose the model incorrectly says the patient is positive for cancer, and the patient is sent for a needless biopsy. The patient experiences some agonizing worry and a possibly painful medical procedure. In the end, however, the patient is relieved that it was just a false alarm and realized the doctor has been prudent.
But suppose the model incorrectly reports the patient is negative and the patient resumes his or her normal life. Six months later the cancer is discovered too late to cure, and after another 4 months the patient dies. The children of the patient suffer the painful loss of a parent in the prime of life. The surviving spouse sues the screening doctor and the sympathetic jury awards 100 million dollars. The doctor’s practice suffers because of the malpractice award, and malpractice premiums go up for all doctors.
Which was worse, the false positive or the false negative? An extreme example if your model isn’t medical perhaps, but surely applicable to many other types of models you may build. Fortunately the NeuroShell Classifier “genetic method” will let you rank right and wrong answers and considers these rankings as it builds the model. In the cancer example, the false negatives will be minimized if at all possible (given data). Of course, in minimizing false positives, it may be necessary to increase false negatives, but this feature is rare in neural net classifiers.
*******************************************************
IV. Thanksgiving holiday closures
November 27th is Thanksgiving here in the USA, so we’ll be closed that day. However, like most companies, we’ll be giving our employees November 28 off as well. Please note that so you don’t get frustrated trying to reach us those days.
*******************************************************
V. Make sure you continue to get the news
We are still surprised that, several months after we sent out our June newsletter announcing release 4.0, users are still calling or emailing to ask when release 4.0 will be available. Seems they all changed their email addresses and never notified us of that. They could have been making money with the new features all of this time. There are many others, no doubt, who still don’t know about release 4.0.
After all these years we are also still getting tech support calls from NeuroShell 2 users who don’t know about the AI Trilogy and its components. So they’re stuck trying to impress their bosses with older technology.
Don’t be left behind – tell us when you change email addresses.
*******************************************************