I. NeuroShell Classifier Model Could Lead to More Effective AIDS Drugs
A neural net model developed at the University of Western Ontario with the NeuroShell Classifier has shown great potential in identifying drug compounds that may be effective against the HIV1 (AIDS) virus. Use of this model could potentially screen out ineffective drugs prior to expensive human testing and insure that drugs that undergo human testing are more likely to be effective. This result is significant because although there are currently over 30 million people infected with AIDS worldwide, less than 10% can afford effective treatment.
The neural net model used the chemical structures of anti-viral drugs to predict the anti-HIV1 activity in living organisms. The study was made feasible due to the existence of a large database of molecular structures of compounds that have been tested for anti-HIV1 activity. The study used the PNN-based genetic training method in order to use all of the data. The genetic training method uses a one hold out training and testing strategy so that the test case was never included in the training set.
The results are published in a paper called “A Hybrid Classification Tree and Artificial Neural Network Model for Predicting the In vitro Response of the Human Immunodeficiency Virus (HIV1) to Anti-Viral Drug Therapy”. Co-authors of the paper include W. R. Danter, D. Gregson, K. A. Ferguson, M. R. Danter and J. Bend of the Departments of Medicine and Pharmacology and Toxicology, University of Western Ontario, London Ontario, Canada. The paper was published in the Proceedings of the ANNNIMAB-1 Conference, Artificial Neural Networks in Medicine and Biology, held in Goteborg, Sweden, from May 13-16, 2000.
II. Adaptive Turboprop 2 is Here!!!!
Many of our NeuroShell Trader customers have asked for a version of our famous Turboprop 2 that retrains itself as often as every bar. Well here it is, formulated as an indicator! Just insert it into your chart or trading strategy just like any other indicator. Never has a neural net been easier to use, yet so powerful!
Adaptive Turboprop 2 (AT2) is the same basic neural network algorithm that is used in the Prediction Wizard of the NeuroShell Trader, Trader Professional, and NeuroShell DayTrader Professional. However, AT2 can retrain itself as frequently as every new bar or after a number of bars which you can specify. The training set is always the most recent bars. AT2 does not make predictions on the training set bars, and its predictions are always “out of sample” (exception: when you specify that you are predicting zero bars ahead). Therefore, there are no walk forward periods to worry about, since the net is essentially walking forward each day (or each bar in the case of the NeuroShell DayTrader Pro).
An AT2 net may be used as an indicator, as an input to a regular Turboprop 2 net in the Prediction Wizard, or it may be used directly in a Trading Strategy. The output is a signal, meaning that the prediction is shown on the current bar, even if it is a prediction for 10 bars from now.
AT2 doesn’t require the optimizer in the Professional versions of the Trader because the parameters are so easy to set. However, when used with the NeuroShell Trader Professional or NeuroShell DayTrader Professional, you can optimize the following parameters:
1. trainbars – the training set size (number of recent bars on which to build the model), up to a maximum of 10,000.
2. aheadbars – the lookahead period (the number of bars into the future the prediction is to be made), i.e., the number of bars ahead that the actual signal is predicted. Be careful not to allow 0 in the range, because then you are predicting the current bar, not the future.
3. actual – the parameter of any indicators used as an output.
4. hiddens – the number of hidden neurons to use when making a network (Turboprop2 in the prediction wizard automatically decides how many to use, but AT2 lets you or the optimizer decide). You can use up to a maximum of 20 hidden neurons, more than enough for this adaptive version.
5. retrain – this parameter specifies in bars how often the net is trained. If retrain = 1, a new net is created for each new bar encountered. If retrain = 10, the net is retrained every 10 bars.
6. input1, input2, etc. – these parameters are the network inputs. You can optimize all parameters of indicators used as these inputs.
AT2 nets allow 2 to 14 inputs. The names of the indicators are correspondingly Tprop2, Tprop3,…, Tprop14.
Adaptive Turboprop 2 is just $249. Call or email today if you have questions.
III. Trader Neural Net Addons Summary Matrix
Now that we have no less than 3 neural net addons for the NeuroShell Trader, Trader Pro, and DayTrader Pro, we can appreciate how the options may confuse you (even WE have trouble keeping these sorted out!) We’re sorry for that, but we figure it’s better to cause some confusion than not to innovate!
Anyway, we’ve prepared a summary matrix so you can see at a glance the differences between our neural net offerings for the Trader. This will also be posted on our web site (www.neuroshell.com) in case it is difficult to read in this email (look under the Addons sidebar).
To view this matrix properly, make sure the window you are viewing it in is maximized as large as your monitor screen.
Here are the abbreviations we’ve used in our headings:
|
Trader –
|
NeuroShell Trader, Trader Professional, or DayTrader Professional |
|
ANI –
|
Adaptive Net Indicators |
|
NI –
|
Neural Indicators |
|
AT2 –
|
Adaptive Turpoprop 2 |
Other abbreviations used:
|
GA –
|
genetic algorithm in the Trader Professional or DayTrader Professional |
|
NN –
|
neural net |
|
GRNN –
|
general regression neural net |
|
PNN –
|
probabilistic neural net |
|
Pro –
|
NeuroShell Trader Professional or DayTrader Professional |
|
PM –
|
parameters |
|
NL –
|
no known limit, but depends on processor speed and RAM |
|
N/A –
|
not applicable |
|
any –
|
the user may select his own value |
|
*this will change in the next release of ANI, when each input can include many lags of itself.
**recurrent net type only
***sparse nets can have 8, 10, or 12; others are limited to 6
Ok, we know the set of available features of our addons is complicated. That is why ANI, NI, and AT2 are optional addons – so we don’t complicate the product itself, yet we can still offer a great deal of neural net options for those who want them! Give us a call if you are still confused about these offerings.
On the other hand, remember: you don’t really have to fully understand these indicators to use them, especially with our optimizer. After all, many people feel pretty comfortable using indicators like RSI and CCI in nets and trading strategies, but we wager that not all of those users understand exactly how the RSI and the CCI work!
IV. Omega Research Intraday Interface
Some of our users may have noticed on www.ward.net that we are beta testing a TradeStation interface for the NeuroShell DayTrader Professional.
Even before the NeuroShell DayTrader Professional was introduced, it was our intent to be able to access the TradeStation 2000 Global Server to obtain intraday data, just as we already can obtain daily bars from it. However, it has not been an easy process due to problems in the Global Server. Fixing these problems has understandably not been Omega’s highest priority, since problems which affect their end users should no doubt be fixed before problems which affect developers like us.
Now, however, we have found work-arounds and are close enough to have a beta test in the field. We still are not promising that the interface we have built will be released, but we are hopeful we can solve problems the beta testers find.
By the way, if you’d like to test it too, you can if you are already a NeuroShell DayTrader Professional owner. You must also have TradeStation 2000 with an intraday feed supported by TS2000. Email us at support@wardsystems.com.
V. Release 2.0 of Advanced Indicator Set 1
We have now finished release 2.0 of Advanced Indicator Set 1. If you are not familiar with Advanced Indicator Set 1, please see the “Addons” sidebar on www.neuroshell.com.
In release 2.0, we are adding the following new indicators:
Hurst Exponent – The Hurst Exponent is useful for determining how random a price time series is. If the series is very random, then in theory it will be harder to predict or trade using standard indicators.
Hurst Significance – Like the Hurst Exponent, the Hurst Significance indicator determines the significance of the difference between the investigated time series and a true random process. Use this indicator instead of Hurst Exponent when the time series has less than 10,000 bars.
Fractal Dimension – The Fractal Dimension indicator is a measure of the complexity of the time series. The complexity number it generates is an indicator of how many independent variables are necessary to simulate the time series, although it is not known what these variables are.
Hodrick-Prescott Window – The Hodrick-Prescott Window indicator computes the Hodrick-Prescott Filter indicator in the last Period bars (including current bar) and draws the filter output on the current bar. Unlike Hodrick-Prescott Filter, the Hodrick-Prescott Window indicator can be used in trading strategies or neural nets used in trading strategies. New data bars DO NOT cause the trend line to change the trend through past data.
Release 2.0 will be emailed free to all requesting and currently registered users of Advanced Indicator Set 1. Email your request, along with your name and serial number to sales@wardsystems.com. Other users may purchase it for $249 plus shipping.