Greetings from NeuroDimension!

The World Leader in Neural Network Software

 

This issue of the TradingSolutions newsletter highlights trading solutions and ways of selecting appropriate inputs for your predictions.

 

In this issue you’ll find:

 

Feature Spotlight

  *  Trading Solutions

        - Using a Trading Solution

        - Tips for Using Trading Solutions

        - Sample Trading Solutions

 

TradingSolutions Tip Box

  *  Choose your Inputs Wisely

 

Products and Events of Interest

  *  Neural Network Training Course

  *  Neural 101 Website

 

Note: You are receiving this newsletter because you requested to stay informed concerning information about TradingSolutions from NeuroDimension. If you would like to stop receiving these newsletters, please see the bottom of this newsletter for removal instructions.

 

=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=

Feature Spotlight

This Issue: Trading Solutions

 

One of the most powerful timesaving features in TradingSolutions is “trading solutions”. Trading solutions allow you to bundle multiple field definitions together into a single concept that can be applied to new data. In addition, they can also be exported and shared with other TradingSolutions users.

 

Using a Trading Solution

Here is an example of when you would use a trading solution. Let’s say you have predicted the optimal signal using the closing price, a moving average of the closing price, and the momentum of the closing price as the inputs. You like the results and would like to try this method against other selected stocks. A trading solution can be used to bundle all of these field definitions together so that they can be applied to new data.

 

To create your trading solution, select “Define Trading Solutions…” from the Tools menu or the toolbar. Press the “New…” button from beneath the list of trading solutions and follow the step-by-step wizard. The key step in the wizard is selecting a source field on which to base your trading solution. In this case, you would select the predicted optimal signal. Once created, your trading solution will automatically include all of the fields needed to create this source field, including the calculated inputs and the desired optimal signal.

 

To use your created trading solution, select “Add New Field…” from the context menu for the data series in which you would like to create the fields. Then, select “Use a complete trading solution to produce one or more values” and follow the step-by-step wizard. All of the required fields will be created and calculated in the new data series. These fields can then be used just like any other fields you may create.

 

Tips for Using Trading Solutions

Here are a few tips to help you get the most out of trading solutions.

 

Let’s say you would like to try the same method with different price information. For example, you may want to use the Median Price (the average of the high and low prices) instead of the Close for all of your calculations. To do this, simply create a Median Price field and indicate to use it for the value of the Close input when applying your trading solution. Every usage of the Close field will be replaced with the Median Price field. (Note: Be aware this can sometimes lead to unexpected results. For example, if the optimal signal included in the trading solution is based on the Close field, it will be created based on the Median Price.)

 

If you have created a set of signal analysis fields to analyze a predicted signal, you can include them in your trading solution by using one of the signal analysis fields as the source for the trading solution, rather than the predicted signal. To list non-signal fields in the list of available source fields, select “Display non-signals in the list of potential source fields.” Since the signal analysis fields require the predicted field in order to be calculated, the predicted field and all of its inputs will be included automatically.

 

Finally, if you have multiple predictions, it is possible to make a single trading solution containing all of your predictions. Simply create a committee of the predictions you would like to include and use that committee as the source for your trading solution. This will include all of the fields required for all of the predictions to recreate the committee.

 

Sample Trading Solutions

For examples of trading solutions in action, see the Solution Library on the TradingSolutions web site. Simply download the solutions, unzip them, and import them into TradingSolutions using “Import Trading Solutions…” from the File menu. The Solution Library is located at:

http://www.tradingsolutions.com/download/solution.html

 

For more information on trading solutions, see the “Working with Complete Trading Solutions” chapter of the online help.

 

=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=

TradingSolutions Tip Box

This Issue: Choose your Inputs Wisely

 

A common misconception with those new to neural networks is that the more inputs you give a neural network, the more information it has and the better the resulting model will be. This is often not the case. The phrase "garbage in, garbage out" is an adage often used to describe the result of feeding bad inputs into a neural network. If the input data is not relevant to the desired output, the network will have a more difficult time learning the associations between the relevant inputs and the desired output.

 

There are three general approaches we recommend for selecting relevant neural network inputs within TradingSolutions. The first is to use a set of indicators commonly used in traditional technical analysis. For example, the Electronic Arts model on the TradingSolutions web site uses a set of indicators that give the neural network the same information used by chartists who use Japanese candlestick charts.

 

Another approach is to use the correlation analysis feature of TradingSolutions. Correlation Analysis is useful for searching for linear relationships between the desired output and a set of proposed inputs. Specifically, correlation analysis determines if an input and the desired output move in the same direction by similar amounts. Using inputs with high (positive or negative) correlation values will often produce good models. See Tutorial 4, Activity 2 for step-by-step instructions on performing a correlation analysis.

 

Those that have a few hours of extra processing time available may want to use the genetic optimization approach to input selection. After selecting 10-15 input candidates, you let a genetic algorithm train the network several hundred times to find the combination of those inputs that produced a network output that most closely matched the desired output. See Tutorial 2, Task 4, Activity 4 for step-by-step instructions on using genetic optimization for input selection.

 

There is no magic number of inputs that works best in all cases. A network with fewer inputs will be less likely than one with more inputs to overfit ("memorize") the training data. It is worth pointing out that all four of the sample models on the TradingSolutions web site use no more than five inputs.

 

Descriptions of the sample models referenced in this article can be accessed from:

http://www.tradingsolutions.com/products/sample.html

 

=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=

Products and Events of Interest

 

Neural Network Training Course
Want to learn more about neural networks in general? NeuroDimension offers a neural network training course that introduces neural network concepts through the use of our flagship product, NeuroSolutions. A new course offering has been added for May 7-11, 2001 in Orlando, Florida at the Grosvenor Resort, located in the Walt Disney World Resort.

 

Our course format allows both novice and advanced users to find a suitable course. Offered courses include: "Introduction to NeuroSolutions", "Fundamentals of Neural Networks and NeuroSolutions", and "Advanced NeuroSolutions".

 

The courses include a copy of our interactive book, Neural and Adaptive Systems: Fundamentals Through Simulations. We are also happy to work with attendees who would like to use their own data in the sample projects. Sign-up early to receive a 10% early registration discount.

 

For details on this new offering, or to sign-up from the Internet, see http://www.nd.com/course/may_2001.htm

For general ND course information, see http://www.nd.com/course

 

For more information and samples of the interactive book, see http://www.nd.com/products/nsbook.htm

 

Neural 101 Website

Neural 101 is a free educational web site on trading stock and futures markets with neural networks. It includes message areas, links, trading indicators, systems and scans. The message areas include special forums on building trading systems, using neural networks, and even discussing tips, tricks, and ideas for using TradingSolutions.

 

NeuroDimension encourages the use of this free resource to foster discussions between TradingSolutions users. You can visit Neural 101 at http://www.neural101.com. You can go directly to the TradingSolutions section by selecting “TS” from the top menu. The message forum can then be reached by selecting “Message Forum” from the left-hand menu.

 

Have an event or product you would like to announce? Contact us at submissions@nd.com.

 

=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=~=

Comments or Suggestions?

 

We appreciate your feedback! Please send us your comments or suggestions concerning this newsletter, our web site, or part of the NeuroDimension product line. Write to us at: feedback@nd.com

 

Have questions about TradingSolutions? Send your questions to: info@tradingsolutions.com

 

Some mail readers may display an attachment at the bottom of this newsletter. This is typically caused when the mail reader is set to display text only. The original, formatted version is sometimes included as an attachment.

 

If you would prefer not to receive these newsletters or subsequent product updates from NeuroDimension, please reply to this letter with the subject heading changed to the word REMOVE.

 

Thank you again for your support of NeuroDimension products!