Summary of findings

14 views
Skip to first unread message

TraderS

unread,
May 29, 2014, 7:06:29 AM5/29/14
to
This is a summary of findings based on suggested software and methods for removing data mining bias:

- The paper by Walton provides a method that is highly questionable if it works and it could introduce more data mining .

- tssb by Aronson is very interesting but more suitable for those with advanced degree in predictive statistics. Documentation is poor, user interface non-friendly and to get support one must pay a high fee. I stopped testing it

- Kantu could be interesting but it silently  crashed on me constantly generating no errors. It was a bit frustrating testing this software. I manage to get no results at all. Maybe someone else will have better luck. 

- Price Action Lab I did not test  as it requires payment of a fee for getting a demo. The cost of a license is also beyond my budget. If anyone has any test results that would be nice.

I think the best solution is developing automated trading system design software from scratch. I am already in the process of raising some funds for that. I think a very good program could developed in a year with a user-friendly interface. 

Fernando Sanz

unread,
May 29, 2014, 9:48:28 AM5/29/14
to automated-tradi...@googlegroups.com
Hello,

Have you tried Adaptrade Software? http://www.adaptrade.com/ it's similar to Price Action Lab and it has a 15 days trial period.

Regards,
Fernando.


On 29 May 2014 13:06, TraderS <serg...@gmail.com> wrote:
This is a summary of findings based on suggested software and methods for removing data mining bias:

- The paper by Watson provides a method that is highly questionable it works and it could introduce more data mining .


- tssb by Aronson is very interesting but more suitable for those with advanced degree in predictive statistics. Documentation is poor, user interface non-friendly and to get support one must pay a high fee. I stopped testing it

- Kantu could be interesting but it silently  crashed on my constantly generating no errors. It was a bit frustrating testing this software. I manage to get no results at all. Maybe someone else will have better luck. 

- Price Action Lab I did not test  as it requires payment of a fee for getting a demo. The cost of a license is also beyond my budget. If anyone has any test results that would be nice.

I think the best solution is developing automated trading system design software from scratch. I am already in the process of raising some funds for that. I think a very good program coudl developed in a year with a user-friendly interface. 

--
You received this message because you are subscribed to the Google Groups "Automated Trading System Design" group.
To unsubscribe from this group and stop receiving emails from it, send an email to automated-trading-sys...@googlegroups.com.
To post to this group, send email to automated-tradi...@googlegroups.com.
Visit this group at http://groups.google.com/group/automated-trading-system-design.
To view this discussion on the web visit https://groups.google.com/d/msgid/automated-trading-system-design/31010843-e842-46b1-89b5-82aa72b29c34%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Sergio Sokales

unread,
May 31, 2014, 6:46:39 AM5/31/14
to
It appears that Adaptrade is genetic programming based. In price action lab website it is stated that:"Price Action Lab™ (PAL) does not use genetic programming, neural networks, permutations or any other methods employed by some other programs that often produce random and curve-fitted systems ."

Genetic programming adds data-mining bias at each step of the process. Each cycle goes over the same in sample and the bias increases. Forward testing is not enough to ensure validation because it is a part of the same data mining process. I think permutations of not so widely known indicators may have a better chance of getting something of predictive value. The problem here is that the indicators these retails program use have no predictive value as has been shown in many academic studies. TSSB may be more useful for that but I find it hard to use. I may borrow some indicators from that program and design my own algorithm. 





On Thu, May 29, 2014 at 4:48 PM, Fernando Sanz  wrote:
Hello,

Have you tried Adaptrade Software? http://www.adaptrade.com/ it's similar to Price Action Lab and it has a 15 days trial period.

Regards,
Fernando.
Reply all
Reply to author
Forward
0 new messages