Feedback request: online tool for finding strategies

[Colleague]

Hi,

we are working on this online tool http://simplyedge.pushbased.net/index.html
It allows the user to find technical strategies on given dataset.
For that it uses GA and predefined simple price patterns.

Its unfinished, but still doing the job as you may see for yourself.

There is so far not much confirmation if people would really pay for such a tool (once it looks better and more useful). So i am starting to look into this direction.

I hope this may be a good place to find some feedback/opinion, especially:

• Would you pay for such a service?
• What killer features would you welcome?
• How to make it for you more trustful ?

If you wanna use it watch first the video, the web is currently not too user-friendly.

Thx a lot.

 

[Solas0077]

There are programs available. This is nothing new.

Trading System Lab
Price Action Lab
Adaptrade
StrategyQuant

and others

 

[Solas0077]

Also there are problems with this approach:

http://www.priceactionlab.com/Blog/...learning-applied-to-trading-algo-development/

 

[Colleague]

There are programs available. This is nothing new.

Trading System Lab
Price Action Lab
Adaptrade
StrategyQuant

and others

Yes they run on your pc, you need to download them install them and at least Adaptrade costs usually 1500$

i mean they are 90s.

Why not rather something that runs in the cloud, that you can check with your mobile, something that you can easily share with someone and so on. (also cheaper)

 

[Colleague]

Also there are problems with this approach:

http://www.priceactionlab.com/Blog/...learning-applied-to-trading-algo-development/

Yes this is a valid argument,

in our approach we are using statistical significance as fitness function, it is also on the user to use the tool correctly.

 

[Solas0077]

Yes this is a valid argument,

in our approach we are using statistical significance as fitness function, it is also on the user to use the tool correctly.

Statistical significance is bogus when used with machine learning.

https://en.wikipedia.org/wiki/Multiple_comparisons_problem