ChaosHunter by Dennis D. Peterson
The purpose of ChaosHunter is to find a formula that gives the relationship between independent variables and a dependent variable. If your dependent variable is tomorrow’s opening price, you’re probably thinking about some sort of formula to use in trading equities. If your dependent variable is tomorrow’s weather, you’re obviously trying to forecast weather. What your dependent variable is really doesn’t matter because ChaosHunter is an extremely versatile tool.
ChaosHunter is written to accept text files using commas or alternative field separators as data files. It is set up to use US or European conventions. It is flexible, and of course, data using a .csv extension works. Since it is trading data we seek, one way to find it is to open a price chart using NeuroShell Trader and save the data; it is saved as a .csv file. Using the File menu, open the saved data file (Figure 1).
As you can see, it is laid out in a spreadsheet format. At this point, you should be thinking, “What portion of this data do I want to use to create the formula, and what portion do I want to use to run the formula and see how it performs?” Click on “Select ranges.” A window (Figure 1: right half) will appear that lets you pick the rows of the data to be used for “Optimization” versus “Out of sample.” The rows you choose for optimization will be the rows used to create the formula, whereas the out-of-sample rows are the ones you use to see how well it performs.