![MQL5 - Language of trade strategies built-in the MetaTrader 5 client terminal](https://c.mql5.com/i/registerlandings/logo-2.png)
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
Yeah, I think I've got it. The results of the grid with initial randomisation apparently don't need to be repeated exactly. It's enough that the result is stable over some small range.
For example, this is what it looks like:
OPTION 1:
EXPERIENCE 2:
The input data, apart from the initial initialization, which was done in both cases, is the same.
Considering that a few losing trades in a row may dampen optimism in the market, a stable result over a certain period of time is not a reason for euphoria. Again, the problem is not knowing which market conditions are most successful for the trained network. It may be so that a few unsuccessful market conditions will destroy the deposit or damage it badly.
So I think you're playing in a casino.
In a casino, the expectation of winning is statistically less than zero. And for me, it's statistically greater than zero. So, dear Doug, not in a casino. In fact!
Considering that several losing trades in a row may diminish optimism in the market, a stable result over a certain period of time is not a reason for euphoria. Again, the problem is not knowing which market conditions are the most successful for the trained network. It may be the case that a few unsuccessful market conditions will destroy the deposit or ruin it badly.
What are you talking about? Or have you solved the problem of risk-free trading in Forex? Of course not!
I know my inevitable risks. The optimal MM is based on them. What is there to discuss? Of course, our objective is to reduce the risks as much as possible, but we must be aware of their inevitability. Otherwise, we risk resembling fighters with windmills.
In a casino, the expectation of winning is statistically significantly less than zero. And in my case, it's statistically greater than zero. So, dear Doug, not in a casino. In fact!
I don't know, honestly.) I don't play in casinos. The MO could be anything. If you have the evidence base in your hands as to why the MO is exactly that, then of course you own that casino called Forex.:)
In a casino, the expectation of winning is statistically significantly less than zero. And in my case, it's statistically greater than zero. So, dear Doug, not in a casino. In fact!
What are you talking about? Or have you solved the problem of risk-free trading in Forex? Of course not!
I know my inevitable risks. The optimal MM is based on them. What is there to discuss? Of course, our objective is to reduce the risks as much as possible, but we must be aware of their inevitability. Otherwise, we risk resembling fighters with windmills.
Do you realise that if you really have a positive MO you are Soros? You can take a loan from the bank and make money, just by relying on your IO. Or maybe you're already a Soros, maybe I'm just saying all this for nothing?) If so, hats off to you.
Given that a few losing trades in a row may dampen optimism in the market, a stable result over a period of time is not cause for euphoria. Again, the problem is not knowing which market conditions are most successful for the trained network. It may be the case that a few unsuccessful market conditions will destroy the deposit or ruin it badly.
You see, comrade...
I'm not a mathematician, of course, but with my feeble mind I can figure out why your nets are losing. It's not even about the nets, it's about you, apparently. As for my deposit, I appreciate your concern about it. Honestly. But as for the rest...
You and I have different goals. I came to Neutron for training (that's what I do, I fight for my deposit), and you came to measure your fists - go for it, of course, but here's the difference...
Neutron didn't come to you, you came to him...
You see, comrade...
I'm not a mathematician, of course, but with my dull wit I can still figure out why your nets leak. It's not even about the nets, it's about you, apparently. As for my deposit, I appreciate your concern about it. Honestly. But as for the rest...
You and I have different goals. The problem is that we have different objectives. I came to Neutron to study (struggling for my deposit), while you come to show off your skills - kudos to you, of course, but here's the thing ...
Neutron didn't come to you, you came to him.
I don't measure anything.) I'm only saying how I see the situation for myself personally with non-growers.:) So far the result is the same as yours, MO today is positive, tomorrow it may be negative, and damn, as they say why it can become so, this is what I wanted to emphasize with this discussion, because this is the biggest problem.
Do you realise that if you have really positive IR, you are a Soros? You can take a loan from the bank and make money, just by relying on your MO. Or maybe you're already a Soros, maybe I'm just saying that for nothing.) If so, hats off to you.
Yes, but no more than the spread...
So keep the hat on:-)
Surprised!
I have several hundred.
Sorry for the translit - there is no Russian keyboard.
1. i skolko prodolzhaetsa obu4enie?
2. ne fakt, 4to +/-5 optimal graniza dlja vesov
3. ne fakt, 4to back propagation optimal dlja nastrojki seti na forex metodika. Ob etom dazhe nau4nye raboty est'.
4. genetika rulit :)
The answer to this question is not meaningful because the learning time depends on the architecture of the NS. What paralocus and I are discussing now runs on candlesticks and has a degenerate optimal architecture - a single linear neuron, with multiple inputs. It is trained by me in a fraction of a second.
If you try to predict the vertical breakdown of kotier, it's more complicated and I use bilayer nonlinear NS. Again, the optimum of complexity for me specifically is two neurons in the hidden layer. Also, I predict a binary series (+/-1) trained in 16-32 epochs (on the order of a second). Again, in this particular case, anything I can feed to the grid input is countable and has a small number of outcomes. For example, for a NS with two inputs, all the possible inputs I can handle are limited to the following outcomes:
-1 -1
-1+1
+1-1
+1+1
I can of course train my grid on a training vector of optimal length to those four outcomes, but if you think about it... all I can teach it is to sort the most probable outcomes into four piles, which is the most optimal training of the grid in a particular case! Now I don't even need NS, I can do it in 1/1000th of a second with a little code.
So how long does it take for me to train the grid? And what is the optimal method of teaching NS in Forex? - Genetics?
DNA rules!
4. genetika rulit :)
>> Genetica rulit. I can justify it.