Simple Explanation of Reactive Corrective - page 2
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
There has been an enormous collective development thing behind this already. No suggestions needed!
Do you use an EA to trade this strategy?
Yes, definitely use an EA. I trade it live.
I would be delighted to see other software developers come along and use pure logic to create new, highly profitable systems. Just please don't babble about "artificial intelligence" and "genetic algorithms"! Those approaches are wrong-headed and a giant waste.
Someone sent me details recently about systems development platform that uses GA. You give it a timeseries, and twenty mins later, it spits out a perfect curve fitted model of the last 20 years S&P. This thing (which isnt actually finished yet) is available at a reduced cost of $60,000 to early adopters ! (but if you aint got 60K to spend you could just buy the commercial GA engine that the systems based on for $199 and feed it the data from an excel file and probably get the same results)
I have to agree with the sentiment that trying to develop a complete strategy using AI (and in particular GA) is going to get you into a whole heap of trouble, for the obvious reason that its not really strategy development, its abdicating responsibility and logical thought to a technoloy in the hope these things spot tradable patterns, with the inevitable data mining bias.
Using GA for specific functional aspects, is a different kettle of fish altogether, and from my limited experience, quite useful !
You have to keep track of how
much in profits you've booked in relation to the biggest losing trade,
and close that biggest loser when your recently closed winners have
exceed it by a certain amount (I like one-third of one percent of the
account).I have been recently working with EAs that use this type of "accounting." It seems effective. Please post any more info that you got up your sleeve, autoFX.
Someone sent me details recently about systems development platform that uses GA. You give it a timeseries, and twenty mins later, it spits out a perfect curve fitted model of the last 20 years S&P. This thing (which isnt actually finished yet) is available at a reduced cost of $60,000 to early adopters ! (but if you aint got 60K to spend you could just buy the commercial GA engine that the systems based on for $199 and feed it the data from an excel file and probably get the same results)
I have to agree with the sentiment that trying to develop a complete strategy using AI (and in particular GA) is going to get you into a whole heap of trouble, for the obvious reason that its not really strategy development, its abdicating responsibility and logical thought to a technoloy in the hope these things spot tradable patterns, with the inevitable data mining bias.
Using GA for specific functional aspects, is a different kettle of fish altogether, and from my limited experience, quite useful !The developers and purchasers of said system should be magnificently rich by now, yes?
This is the same kind of flimsy stuff I read about all AI and GA stuff...it's not finished yet...but the developers want big bucks for it...it "predicts the past" really well....uh, no, I don't know of anyone making good money with it...sorry....
I have been recently working with EAs that use this type of "accounting." It seems effective. Please post any more info that you got up your sleeve, autoFX.
There is a guy floating around fx forums using the nick "Dr. Zogg" who appears to have reached the same conclusions I have.
My ideas are encaplusated in EAs...I gave a good thumbnail in the first post of this thread. Everything else up my sleeve gets posted on MY forum.
autofx dont u feel like ur cheating on ur forum when u post here? hehe im jk
im gona open an account with ibfx and and join ur forum check it out.
by the way im a noob but i learn quick, regarding AI here is my opinion if it matters at all. AI and neural netorks i think especialy in these type of forums there is alot of misunderstanding and misperceptions about the usage and application of AI. i have an AI EA that places orders via price predictions, and it does alright, but this type of aplication of AI is usless and trash. im intrested in adaptive ai. to make adaptive indicators, or adaptive systems. we can already make adaptive EAs with out the need for AI. but AI can strengthen the adaptive components of a system. the people that make the AI dont know how to trade, the people that use the AI dont know how to trade but someone that does know how to trade and can impliment AI in his own way. will have an edge. but one can not say, oh im all about EA's and using systems on computers. its quicker and they dont get tired. and say AI is a waste of time, thats a contradiction.
I'm not cheating on my forum, nah. I own/admin that forum, so can I cheat on myself?
This is not true.
UNLESS you lump the logic used to implement very simple trading tactics under Artificial Intelligence (AI), which I do not.
However, Expert Advisors (EAs) could be lumped under "Expert Systems", which to some is a form of Artificial Intelligence...
It's all just terminology.
What I'm referring to when I say AI and Genetic Algorithms is a very exotic type of software that is usually overkill and is trying to predict the future, or is being sold as something that can predict the future.
There are many ways to think about trading. I come against the school that says, "You have to be a very wise analyst who understands the technical, fundamental and psychological aspects of a market to trade it profitably." I say, "Bull hockey, I have simple software scripts that can trade profitably, and they are extremely simple...they implement effective trading tactics, don't get tired, are accurate, and react very quickly. They don't try to predict what will happen next, and they don't need to."
It seems this kind of talk gets some folks very riled up. Why? Because they have been making good money collecting commissions and fees from people whom they've convinced they are gurus.
I come against the school that says, "You have to be a very wise analyst who understands the technical, fundamental and psychological aspects of a market to trade it profitably." I say, "Bull hockey, I have simple software scripts that can trade profitably, and they are extremely simple...they implement effective trading tactics, don't get tired, are accurate, and react very quickly. They don't try to predict what will happen next, and they don't need to."
I like the way you think! Now if I only had enough money to join your forum.
What I'm referring to when I say AI and Genetic Algorithms is a very exotic type of software that is usually overkill and is trying to predict the future, or is being sold as something that can predict the future.
GA's and GP's dont try to predict the future, and Ive yet to see anyone (other than complete charlatans) suggesting they can be used to predict the future. Theyre a tool, nothing more, nothing less.
Even MT use's GA's for optimisation, these types of tools are commercially available, low cost and reasonably user friendly.
Im the worlds greatest sceptic when it comes to AI, I'll even go as far as to argue there is no such thing as AI. However I will acknowledge, certain AI techniques are extremely suitable for specific types of problem, and to discount em is foolish (IMHO
).
Clearly anyone who thinks that you can buy a system that automatically develops profitable strategies for 60K needs their heads examing LOL, then again, people think you can buy an EA from eBay for 80 bucks, go figure