A Dynamic Self-Adaptive Trading System

[lindq]

Responses like yours are PRECISELY why ET is suffering lately from a lack of postings from experienced traders.

My comments were obviously meant to help someone who has put a lot of work into a system he is trying to commercialize. But the copy is vague and not well written, and as with any commercial site, he would be well advised to get some help.

He asked for advice and I gave it. If that bothers you, then run off to Motley Fool and post there.

Quote from burnin:



If you took the time to check out his site, youwould have known that he is from China.

Do all of us here a favor, and try to add something positive, instead of being arrogant.

Its guys like you that ruin this forum.

Grow up.

 

[danxl]

Quote from Roscoe:

Is the system intraday only or can it be modified to look at daily data?

IPO, the more signals a system generate, the easier it will be to verify that a system has an edge. Sometimes trading smaller time frame does not mean more profit, because you have to pay more commissions. But with more signals you can get a more robust and more reliable system.

My system is designed to work with tick data, so that daily data, or minute data, would not work.

 

[danxl]

Quote from Hittfeld:

Before I install: Is it restricted in any way??

You don't need to install anything, you just boot from CD, and should see a greeting message and some instructions on the screen.

Quote from Hittfeld:

For other futures symbols what data and how many days/years would you need?

Depends on what symbol. And the symbol must be active enough for intra-day trading.

 

[danxl]

Just setup FTP download for the 31M CD image file.

Here is the download page.
http://www.highland.widge.org/download.html

 

[CalTrader]

Quote from danxl:

Hi, every one, I am back. This is the follow up of the message I posed a year ago here in ET to tell the world that I had developed an Dynamic Self Adaptive Trading System, a trading system that does not require parameters optimization. The framework of the system was done one year ago, but it takes me a long time to make it work with short term day trading, and it was now fully automated and seriously tested with emini s&p data. I offer free trial of my trading system to any one who are interested, and, seriously, I am looking for people who can perform some more testing on it, and eventually, trade it. Please visit my website at www.highland.widge.org more information.

These types of systems really are extremely difficult to design so as to have any practical value. You may get a few people to look at this but my suggestion is to either publish it as open source or to charge a nominal fee for source and design details and try to make your money selling services related to your technology - assuming its worthy of international patent.

Your "system" may indeed process the data feeds in some unique way that truly adds value but without more infomation few people will spend valuable time to evaluate / test your product for you. Most of this reluctance is based upon the very large number of systems writers that claim some type of special edge in their approach to data mining/processing: Nearly all of the claims end up being refuted upon close examination.

 

[danxl]

Quote from trader333:

This is way too technical for windows users. There are no instructions for newbie, it assumes you are linux programmer and know how to program and compile for real time data. Not everybody is programmer here. I think you should include instructions for windows user for realtime data from several data feeds. Believe me more than 90% traders use windows whether they like it or not.

The trading system is developed with the needs of the windows users in mind. All operations, including the backward testing and real time testing, can be done on a windows PC, with an easy to use GUI interface You are not required to know any thing about Linux at all.

However, some more work need to be done for the support of real time IB data feed. I will be working on it right now. It should be done in one or two week or so, I think.

And for the backward testing, the support is already very good, even for a windows user that does not know about Linux at all. You can get started right now.

 

[T-REX]

By the way, my spelling and grammer SUCKS!!! and I'm an American born and raised.

I think what everyone here is trying to say is that you should have more clearer info on the nuts and bolts of your system. It should be clear and concise and written in such a way that the writer (instead of) the reader suffers!

Details. Details. Details. Unless you can explain and divulge of your info in a straight forward maner....no one will be able to decifer its contents.

Here is an example of my peformance record of my DayTrader System useing Daily bars with a starting capital of $3,000 U.S. dollars trading 2 contracts.

The report shows buy & hold verses the day trade results ameriterized over an eight year period. (1995 - 2003).

Hope this helps.

p.s. I happen to like your site.

 

[danxl]

Can anyone help to convert some historical data from xpo and omz format to ASCII format? I need to do some more backward test with the data before 2001, while most data I already tested is in 2002. I found the data I need at http://www.medianline.com/ But I can not handle the file format.

 

[yoyo2000]

No further discussion?
I can't access the address......anyone share the package which downloaded from there?
Thanks

 

[kut2k2]

Quote from yoyo2000:

No further discussion?
I can't access the address......anyone share the package which downloaded from there?
Thanks

Don't waste your time. The system went away for a reason. The guy was either a charlatan or delusional:

Quote from danxl:

Hi, every one, I ... developed an Dynamic Self Adaptive Trading System, a trading system that does not require parameters optimization.

Quote from vanilla2:

can you describe more about what this system is adapting to? it sounds like it's continually re-optimizing some indicator on a continual basis. is this correct?

Quote from danxl:

You say re-optimizing, this might be one way to build an adaptive trading system, but it is really hard to do it efficiently. My trading system is not based on "continually re-optimizing", etc. It is more efficient than that.

"More efficient than real-time optimization."