Australian (ASX) Stock Market Forum

StrategyQuant

Joined
30 June 2007
Posts
7,200
Reactions
1,225
Currently doing a trial. Only takes an hour to get the hang of it, if you feel like having a go.

It appears to do everything claimed on the web page (that's unusual for trading software). It crunches a massive number of quotes/operands/integers/indicators and spits out an optimized equation in various formats. I've been mucking around for about 5 hours and so far no "huh??" moments, which is also unusual when testing new software.

It is more comprehensive than Chaos Hunter, and much better equipped to test for robustness on unseen data. I feel very confident curve fitting can be minimized, however finding a good result could take.... how long have you got?! I feel like I'd need a supercomputer to do this justice. It's clear a lot of work and care has gone into creating this product but I'm left wondering if it could actually make me money.

Would like to see what sort of results others get. You can import your own data or use the data they provide.
 
Currently doing a trial. Only takes an hour to get the hang of it, if you feel like having a go.

It appears to do everything claimed on the web page (that's unusual for trading software). It crunches a massive number of quotes/operands/integers/indicators and spits out an optimized equation in various formats. I've been mucking around for about 5 hours and so far no "huh??" moments, which is also unusual when testing new software.

It is more comprehensive than Chaos Hunter, and much better equipped to test for robustness on unseen data. I feel very confident curve fitting can be minimized, however finding a good result could take.... how long have you got?! I feel like I'd need a supercomputer to do this justice. It's clear a lot of work and care has gone into creating this product but I'm left wondering if it could actually make me money.

Would like to see what sort of results others get. You can import your own data or use the data they provide.

A bit like Bryants adapt trade builder....genetic algorithm?
 
A bit like Bryants adapt trade builder....genetic algorithm?

Looks similar yep. Probably good if you have an entire bank of super-fast, multi-threaded linked computers and can run millions of queries at once. Or if you know someone at a university who can get you access to a supercomputer. Discretionary remains the ultimate aim for me, but I like to look at everything.
 
Russian

Random selection of components
You have no idea what's inbuilt in the algos
I'd say there's a really good chance that look ahead
Formulas are built in to data testing so you'll
Always find a profitable system on old data.
 
Random selection of components - no, you select input components yourself
You have no idea what's inbuilt in the algos - actually, you know exactly what the formula is
I'd say there's a really good chance that look ahead - no it goes to great lengths to avoid this
Formulas are built in to data testing so you'll
Always find a profitable system on old data. - not my experience - very hard to find anything you'd trade
 
Looks similar yep. Probably good if you have an entire bank of super-fast, multi-threaded linked computers and can run millions of queries at once. Or if you know someone at a university who can get you access to a supercomputer. Discretionary remains the ultimate aim for me, but I like to look at everything.

Not sure if this can be done with a Windows Shell, but Amazons AWS has super computer for rent. I've seen them do some crazy things.
 
Random selection of parameters and configuration of inputs. The inputs themselves are chosen by the user. Random is good, by the way.
 
Currently doing a trial. Only takes an hour to get the hang of it, if you feel like having a go.

It appears to do everything claimed on the web page (that's unusual for trading software). It crunches a massive number of quotes/operands/integers/indicators and spits out an optimized equation in various formats. I've been mucking around for about 5 hours and so far no "huh??" moments, which is also unusual when testing new software.

It is more comprehensive than Chaos Hunter, and much better equipped to test for robustness on unseen data. I feel very confident curve fitting can be minimized, however finding a good result could take.... how long have you got?! I feel like I'd need a supercomputer to do this justice. It's clear a lot of work and care has gone into creating this product but I'm left wondering if it could actually make me money.

Would like to see what sort of results others get. You can import your own data or use the data they provide.

I'll have a bat at it. I like the trading bots because I look to see what it is they are actually doing and try to see if it will work with my trading. So far everything I've tested are trend following systems. Works, but don't need a bot for that.
 
Not sure if this can be done with a Windows Shell, but Amazons AWS has super computer for rent. I've seen them do some crazy things.

I don't think I'll pursue that option myself, but did you happen to follow up?

@Roller - I think AB could do something very similar if you had the coding ability. But it's not set up to 'find' equations.
 
Top