Looking for patterns - page 6
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
Vitaly, a broad question.
It is impossible to create a completely profitable EA, or to trade manually in the same way. The ideal EA, which everyone strives for, makes losing trades. But the total profit is greater than the total loss, and this condition should be met at any time. There is enough history to find the patterns today, so I think there is no point in creating a self-training robot. Another thing is that the Expert Advisor should have several strategies for working in different parts of the market. Therefore, we need a mechanism for recognizing such periods in order to switch between strategies, because the price changes from time to time.
And constant self-learning smacks of adjusting parameters to a small section of history. With disastrous results in reality. Personal opinion, with no claim to objectivity.
The fact is, Alexey, that digital technology has made a huge leap in recent years. Creating a completely profitable Expert Advisor is not an impossible task if you apply the necessary resources. Imagine an AI that is connected to a multitude of modules containing databases of billions of patterns, continuous interactive communication with the publication of formal events, including details of everyday news that can affect the quotes. That is, a kind of mini Skynet like in Terminator (or better even Google), in real time determining reactions based on a set of events run through patterns, etc...
It's not all that impossible and even more so, if you scroll through similar digital developments and information on AI, a lot more than what I'm pondering here has already been implemented.
But the question is also who benefits from creating such an automated system? After all, it is certainly least beneficial to the ultimate beneficiaries of the current structure. If this is implemented, there is no point in exchanges at all. It is a difficult and painful question which there is no point in elaborating on in detail.
But mainly, I have received an answer to my question. Thank you.
P.S. In my opinion, the existing automata exactly as you expressed it, cannot cover many things. And humans have a psyche and nervous system that often work against themselves. In this regard, robots could ideally replace humans, but not in their current state. Again, based on the goals of the field, the availability of such AI to any of the outside parties, promises the collapse of the whole system.
Statistics of the size of a useful wave movement on the EURUSD history, in pips of a 4-digit quote (multiply by 10 if necessary).
The figure demonstrating what is a useful movement is on the previous page.
The same data in probabilistic form:
What does it mean?
The price in each wave does not pass through more than
-120 pips 90% of the time,
- 60 pips 70% of the time,
- 35 pips 50% of the time.
Statistics of the size of the useful wave movement on the EURUSD history, in pips of a 4-digit quote. (multiply by 10 if necessary)
Picture showing what is a useful movement on the previous page.
What does it mean?
the price in each wave does not go over
-120 pips in 90% of cases,
- 60 pips in 70% of cases,
- 35 pips 50% of the time.
It turns out that the most productive and correct option is scalping on accounts with minimal spread.
Imagine an AI that is linked to many modules,
I don't think there are any such robots. The problem with the AI is its large number of weighting factors. Three, five, ten parameters can't adjust to a long story. Hundreds can. And on history you get a smooth upward almost straight line increase in the deposit, but in real trading it's a disaster.
It turns out that the most productive and correct option is scalping on accounts with minimal spreads.
Well look, 18% of the time the price does not pass more than 10 pips and turns around. That is, if you expect more movement, there will be a loss every 5 times.
I don't think there are any such robots. The problem with AI is its large number of weighting factors. Three, five, ten parameters can't adjust to a long story. Hundreds can. And on history you get an even upward almost straight line increase in the deposit, but in real trading it's a disaster.
And you just pay attention to Google, which is capable of instantly translating text into any language and reformatting it into voice and vice versa. It can also read text in pictures, identify objects in pictures, even determine a person's identity 100% by face, fingerprint and other parameters. At the same time, its associated modules in general are capable of controlling almost anything, as long as it is network-linked. These capabilities are already far beyond what a fully profitable vending machine could be. A vending machine requires orders of magnitude less data, processes and resources. If the developers of corporations like Google wanted it, they could easily make such a system out of their modules.
Look, 18% of the time the price doesn't go more than 10 pips and turns around. So if you expect a larger move, every 5 times there will be a loss.
That is correct. But you take into account the ratio of the movement history, but you don't take into account the determinants of those movements. If you analyze this particular pair, you will probably catch the moment when you can take not only 10, but 50 pips. And most likely you will not open a position when there are no clear-cut preconditions.
Alexei, what tools do you use to calculate such data?
Created a zig-zag type of script.
And you just pay attention to Google,
These are different elements. Real life datasets for learning and the uncertainty of the future.
Created a zig-zag type script.
These are different elements. The data sets for learning from real life and the uncertainty of the future.
What is the uncertainty of the future if you have used your script to define the boundaries of the movement? This is already a huge achievement. I am, to be honest, impressed by your analysis. And the current interactive systems have fewer holes than in the minds of their developers. Don't draw blind conclusions. You're the developer. Just assess the scale of what has already been achieved. Really compare. And realise that these features have already been overshadowed. They just aren't needed.
And the future is getting dumber by the day.