I. How to keep receiving this newsletter
It’s really simple: just tell us when your email address changes! If you like this newsletter, remember to call or email to tell us if you change your email address. Otherwise, you’ll stop hearing from us.
II. Student Wins Science Fair with Sales Prediction Project
A high school student swept the honors in the Computer Science category of the Frederick County, Maryland, 2001 Science Fair. The judges awarded the project “Paging Neural Nets, Company Statistician” with first place in the category, a U.S. Army recognition award, a U.S. Air Force Award, and an Intel Excellence in Computer Science Award of $200.
According to the student “I presented the project to about seven judges. Many of the judges have worked with neural nets before, and they seemed to like my project very much.”
The project used both the NeuroShell Predictor and regression analysis to do sales predictions for the student’s web-site business. The business, http://www.virtue.nu/moonslush, caters to teenage girls in need of affordable lip gloss and other teenage accessories.
The following quotes were taken from the student’s project:
“How accurate can a neural network predict sales for my Internet start-up Moon*Slush? What independent variables most influence the amount of business generated, according to the neural network? How do the predictions for the non-linear neural net compare to the predictions of a regression analysis?”
“I was able to get an optimal correlation of 0.844761, and, in general, variables pertaining to the events and conditions of the days are the best variables to use to create the best neural network for sales prediction. These type of variables can include the season of the year, weather (temperature, precipitation, pollen count, etc.), the number of holidays, distinctions between days of lesser stress, consideration of outside, monumental events to the target demography (concerts, sport events, Academy awards, season finales, etc.), events at other websites, and so on.”
The student also experimented with other variables such as colors in the website, advertising, and number of items for sale.
“Compared to linear regression, neural networks have a stronger accuracy if the data is non-linear; simple linear regression models do not take multiple variables into account. A neural network also does not have to predict in a linear path; the linear regression models are restricted to only predicting in a linear route, which therefore increases the amount of error if the data is non-linear.”
“From my experience with neural networks, other small companies may find neural networks as a great solution for sales prediction and company optimization. This method is an affordable and easy way to predict sales without the hassle of having to hire a full-time statistician. Neural networks also forces an entrepreneur to discover information about his business; the knowledge that is gained from the use of neural nets to predict sales is priceless, as it can dynamically improve one’s company for years to come.”
III. Price Increase
The bad news: we are announcing a price increase on all NeuroShell Trader add-ons, effective June 1, 2001. After May 31, the add-ons we sell will cost $299 plus shipping instead of $249 plus shipping. We are forced to take this step because of the higher costs of producing and shipping CDs, both in labor and materials. This is our first price increase in several years.
The good news: if you are willing to take delivery of new add-ons by Internet download, and make your own backup copies on CD, Zip Disk, or other backup media, you will receive a $50 discount on your purchase. Not only that, but you will not have to pay any shipping costs either, so the real discount is even more than $50! The addons are about 850KB and 1200KB in size (exception: the Data Exchange Add-on is about 2MB).
So if you were thinking of purchasing an add-on, and you have to have it on CD, now is the time to get it! Replacement CDs will also be $50 after May 31.
Add-ons affected are:
Adaptive Net Indicators – neural nets especially adapted to pattern recognition, some of which automatically include lags of inputs. Useful for building your own adaptive moving averages.
Adaptive Turboprop 2 – like the neural nets in the Prediction Wizard, but they operate like an indicator and automatically retrain themselves. You can optimize training set size, walkforward interval, number of hidden neurons, and even the lookahead period.
Neural Indicators – classification neural nets that give you the probability that the current situation is a buy or sell opportunity. These nets don’t predict any price, they give you a direct signal probability. Some of them are “recurrent nets” that automatically look back in time.
Advanced Indicator Set 1 – a set of indicators that users requested including chaos indicators, an indicator that gives the phase of the moon, and curvefitting trendline indicators.
Cluster Indicators – like the neural indicators, these give a buy or sell signal (the probability that your issue is a buy or sell opportunity), but they work by clustering instead of using a neural net.
Fuzzy Pattern Recognizer – Lets you describe patterns in which you are interested, and then tells you when these patterns appear. Uses fuzzy logic to find the patterns, so it can tell you when a pattern is “like” the one you are seeking.
Data Exchange add-on (DataX) – lets you communicate with the DayTrader via Visual Basic or C++ programs instead of writing DLLs.
IV. NeuroShell Trader 3.21 problems
A dozen or so users of our NeuroShell Trader have had problems with the new release 3.21. Most of these involved installation problems, where a DLL or OCX file failed to properly replace an older one. A few involved problems interfacing with the Omega Global Server. Our programmers are working hard to resolve all such problems. The registry on the Windows operating system has turned out to be even more of a nightmare for Windows users than the old “DLL Hell” was! Some of these problems are no doubt due to corruption in registries, but we may find bugs or better ways to install this now very large system. There are some other minor bugs, so we will no doubt have a maintenance release 3.22 in the coming weeks.
V. We’ll be closed Friday the 13th
Ward Systems Group offices will be closed Friday, April 13th, 2001. We’ll reopen as usual on Monday the 16th.
VI. Live from the forum
We’ve gotten lots of thanks for building the User Forum on www.ward.net, because many great ideas have been discussed there. Techniques have been shared, and even source code and template files have been donated. (We’re talking about the Trader section of course. We’ll be removing the other product forums soon for lack of use!)
We get a great deal of tech support calls here that start out something like this: “What do most of your other users do about X?” Unfortunately, we probably have little knowledge about what our users do with regard to any given X. So usually, we can only tell you what WE suggest with regard to X. However, we’re the first to admit that a great number of our users may have ideas about X that never occurred to us.
So next time you wonder what others are doing, why don’t you ask it on the forum? And while you’re there, why don’t you post your ideas about what others are asking? It is a two way street. The forum will be an invaluable tool as more of you post there. Please don’t just “lurk”, thinking you have nothing to add. Chances are, your ideas are as good as anyone else’s. Besides, most of the time, there’s no “right” answer anyway. It’s good to see lots of different thoughts, and then people can make up their own minds about which ideas they’d like to pursue. No thought is too dumb, and we’d like to see more thoughts there.