Home PageJRC Logo




colorbar

BrainCel

Neural Net Module for Microsoft Excel

colorbar


AmiBroker
eSignal
Extreme Charts
Generic DLL
Investor R/T
Market Delta
MATLAB
MetaStock
Microsoft Excel
MultiCharts
NeoTicker
NeuroShell
NinjaTrader
Tradecision
Trade Navigator
Trading Solutions
TradeStation

Create leading indicators

Create complex concurrent indicators

Apply to stocks, futures, ... almost anything



Want to create a formula that responds to data a certain way, but have no idea how? Here's a way to create them without performing any calculations yourself.

Neural networks can be "trained" to analyze your data and automatically manipulate hidden formulas to approximate your desired responses.







Why Neural Networks?


Neural networks are mathematical functions that you "train" to covert input data to whatever you specify as output data. For example, you may want to convert various technical indicators into:

You may not know the math to perform such calculations, but a Neural Net may be able to figure it out! Afterwards, you can develop trading decisions based on the neural net's forecasts.

The power of Braincel is well established. In each of the following articles in Future's Magazine, Braincel was used to build a leading indicator...





Braincel in Microsoft Excel


BRAINCEL is an Excel add-in that enhances your forecasts with the power of neural networks. Amazingly, no knowledge of neural net math or statistics is required to use it. You supply both input data and desired output (target) data in rows and columns on a spreadsheet. Braincel will study this data and after training, you supply it with new input data and Braincel will give you answers that will astound you! It's almost as simple to use as Excel's standard regression tool, but more powerful!

To create a neural net, you simply set up your data on an Excel spreadsheet, select a few options, and the neural net will automatically train and test itself. The neural net models you produce will work in ...

Novices enjoy the guidance from standard dialog boxes and programmers love the power in using Excel's VBA to automate the more complex tasks that Braincel can perform.

Braincel's advanced features include . . .





Braincel in TradeStation


Here's how to create a neural net for use in TradeStation 2000i, 6, 7 or 8...

  1. Plot your favorite technical indicators on your TradeStation screen.
  2. Export the chart data to a .CSV file and open the file with Microsoft Excel.
  3. Create and train a neural net on the spreadsheet data and save it.
  4. Write code in Easy Language to apply the neural net to market data. You can access your neural net within INDICATORS, SYSTEMS, and USER FUNCTIONS. (Example Easy Language code is provided.)
  5. Run your new studies on real-time or end-of-day data!

For a working example of how you can access from TradeStation Easy Language, a neural net built by Braincel, look for "Braincel - Neural Net Example" in our "Free Stuff" section.








Software Requirements


To run Braincel and create neural nets, you need one of the following:

Microsoft EXCEL 97, 2000, 2002 running on Windows 2000, XP, Vista
Microsoft EXCEL 2003 running on Windows 2000, XP

You can run your newly created neural nets on market data in Microsoft EXCEL and TradeStation


LATEST VERSION: Braincel 3.7released JUNE 2005

To determine if you have the latest version, proceed as follows:

  1. Open the folder on your hard drive containing the installed Braincel files
  2. Right click on the file "BRAINCEL.XLL" and select PROPERTIES
  3. For version 3.7, the "Created" date should read 27-MAY-2005.
  4. To order the latest version, go to our SHOPPING CART and select the "Upgrade/Replace" department.




Additional Information


Download our TECHNICAL REPORTS. Lots of charts and comparisons.

Read our CUSTOMER's LETTERS. See what others have to say.

Read our FREQUENTLY ASKED QUESTIONS. For answers to technical issues.

How to order Prices, order form, specific questions to ask.


SPECIAL REPORT: BACK-PERCOLATION

Braincel uses a newer algorithm called Back-Percolation instead of the rather old formula called Back-Propagation. Back-Percolation was designed by Mark Jurik and this report explains his philosophy of the successful algorithm. For a real visual treat, the report displays parameter learning trajectories of amazing symmetry and beauty. --- Go to TECHNICAL REPORTS page.














[ Copyright ] [ WebMaster ]