Apologies err, the word "not" did not make it in my post - making my point difficult to understand - my apologies. You can guess where the word goes in the above post. My point: Let good technoology take the grunt work out of the equation and allow your decisions to create greater impact. You cant just dismiss tools and technology b/c successful traders dont use it. A better generalization might be: Evaluate tools and arrive at your own conclusions.
I think down the track we will see NNS used as a way of polishing trading systems. The NN could monitor the performance of a number of systems along with certain market characteristics ie volatility etc then choose the systems and money management rules that best fit the underyling conditions. I still think you would need profitable systems in the first place though. This of course would be more the territory of serious hedge funds than private traders due to the costs and resourses such a system would consume. Runningbear
Thanks again to all who responded. EQ, how did (do?) you use JMA in your system? Jurik called it "optimal noise reduction" but it really is sub-optimal in the sense that you still have to determine the best length parameter (let's pretend the so-called "phase" parameter is set to its default level) based on some (user-specified?) criterion. Does Jurik give guidelines on how to do this, or are you just left hanging with a 'cool tool' and the same old trial-and-error method of plodding towards optimization? TIA.