Using artificial intelligence at MTS - page 5

New comment
 
Mathemat:

Reshetov, you have to keep it simple. Naturally, we are only talking about estimating probability by frequency. That's what statistics is for - otherwise a theorist would be a completely useless abstraction. The variance, for example, we know how to estimate from a limited sample. And we know how to estimate the chances that the sample estimate differs from the population estimate by no more than a given value...

Then such an indicator should be given confidence intervals as input parameters and the probability that these values will come out or stay within these same intervals can be obtained by frequency.
 

I doubt that there will ever be an oscillator that accurately calculates probabilities. Even statistics can only reveal the frequency of events in a sample. But the value of frequency tends to the value of probability only in samples in which the number of investigated events tends to infinity. The Law of Numbers says that the difference between frequency and probability tends to be as small as infinity in a sample. It is possible to calculate the probability of obtaining an error in the frequency and number of events under investigation. But there is no method yet devised to determine the probability of an event itself, other than waiting for an infinite number of events. Except for the case when it is known, but then there is nothing to calculate, because in this case the amount of information is 0.

Frankly speaking, it is funny to read such deep speculations about the nature of probability - and nothing in essence. If you are an expert in mathematical statistics, maybe you can say something specific, for example, to give such a definition of an event, so that for each indicator value you can calculate the frequency of its occurrence on the history. Or in some other way to formulate a correct statement of the problem.

I'm sure for some reason that even such "experimental" value of the probability of one or another price reversal in the future will be not less profitable, than your artificial intellect. Especially if we consider the fact that the topology of the area of three parameters values, that actually provide the profitable trading, can be much more complex than just a half-space and the linear filter used here is not supported by anything.
  
double perceptron() {
   double w1 = x1 - 100;
   double w2 = x2 - 100;
   double w3 = x3 - 100;
   double w4 = x4 - 100;
   double a1 = MathSin(Time[0]);
   double a2 = MathSin(Time[1]);
   double a3 = MathSin(Time[2]);
   double a4 = MathSin(Time[3]);
   return (w1 * a1 + w2 * a2 + w3 * a3 + w4 * a4);
}


There! Variables x1-x4 are sufficiently optimised in steps of 10 (0-100).

 
Reshetov писал (а):
But you didn't go to school, did you pick up cigarette butts at bus stops during school hours? Now you pretend to be a hoser and pretend to be an "expert" on line filters. When in fact you're a lamer and a two-timer. And all the same sooner or later it would be found out, because being illiterate you will make a mistake on some leap. And once they figure out your bullshit nature, don't expect any leniency from the people around you. Your opinion will be ignored.


Yuri, why so rude? Leninist style somehow argue: instead of convincing you of what is right, my personality you pounce on, calling me names. You don't know me. My knowledge is all in order: I got gold medal from school 25 years ago, obtained red diploma from institute (Moscow Electrotechnical Institute, by the way, if somebody has heard), Ph.D. in Electronics, I have 11 patents here and abroad. If you're going to scold me and call me names again you'd better not. Save that talent of yours for the bazaar.
 
gpwr:
Reshetov:
But you didn't go to school, did you pick up cigarette butts at bus stops during school hours? Now you pretend to be a hoser and pretend to be an "expert" on line filters. When in fact you're a lamer and a two-timer. And all the same sooner or later it would be found out, because being illiterate you will make a mistake on some leap. And once they figure out your bullshit nature, don't expect any leniency from the people around you. Your opinion will be ignored.


Yuri, why are you being so rude? It's a Leninist argument: instead of convincing me that you're right, you attack my personality, calling me names. You don't know me. My knowledge is all in order: I got gold medal from school 25 years ago, obtained red diploma from institute (Moscow Electrotechnical Institute, by the way, if somebody has heard), Ph.D. in Electronics, I have 11 patents here and abroad. If you're going to scold me and call me names again you'd better not. Save that talent for the bazaar.
Where in the alley or in the subway did you buy your diploma and dissertation? Doctors and inventors who do not know high school mathematics are certainly not uncommon these days because almost everything is bought and sold.

And I'm not attacking personalities, I'm attacking competence, or rather the lack thereof. I did not make you imposter yourself as if you were an expert in any field. Well, not only did you fart in the puddle in that very field, you also attribute the title of doctor of science to yourself. Go learn the material, you two-timing bastard.
 
Integer:
double perceptron() {
   double w1 = x1 - 100;
   double w2 = x2 - 100;
   double w3 = x3 - 100;
   double w4 = x4 - 100;
   double a1 = MathSin(Time[0]);
   double a2 = MathSin(Time[1]);
   double a3 = MathSin(Time[2]);
   double a4 = MathSin(Time[3]);
   return (w1 * a1 + w2 * a2 + w3 * a3 + w4 * a4);
}


There! Variables x1-x4 are sufficiently optimised in steps of 10 (0-100).

The sine arguments can be values from 0 to 2 * PI, not taken from the ceiling and sucked out of 21 fingers like yours.
  
double perceptron() 
  {
   double w1 = x1 - 100;
   double w2 = x2 - 100;
   double w3 = x3 - 100;
   double w4 = x4 - 100;
   double a1 = iAC(Symbol(), 0, 0);       //       Вот это место
   double a2 = iAC(Symbol(), 0, 7);
   double a3 = iAC(Symbol(), 0, 14);
   double a4 = iAC(Symbol(), 0, 21);
   return(w1 * a1 + w2 * a2 + w3 * a3 + w4 * a4);
  }

Good evening. I would like to welcome all participants of this interesting topic. Dear Reshetov, please explain the following point: in your Expert Advisor code, where the Perseptron function is calculated, AC value from zero bar is used. This means that the expert looks into the future while testing since it uses the current value of AC that hasn't yet been formed in reality. This calls into question the objectivity of testing and the results of forward testing on the remaining history.
If I am wrong, please explain this point in more detail.

Sincerely, Pooh.
 

2 gpwr

"Don't cast your pearls before swine, lest they be kicked around".

I just wonder how long you've been arguing with a Boy Scout who never grew out of his trousers.
He wrote four lines of code, called it a neural network, knows the plane's uranium.
I would have given up hope to hear anything clever from him long ago.

But then, sure, let him be rude. While the moderator is asleep.

PS Integer, the example is great ! But he hasn't understood anything. :-)))

 
Pyh wrote:

In your Expert Advisor's code, where the perseptron function is calculated, the AC value from the zero bar is used. This means that the Expert Advisor is looking into the future when testing, because it uses the current value of AC, which has not really formed yet. This casts doubt on the objectivity of testing and the results of forward tests on the remaining history.


Pyh, he does not look into the future. The testing mode is by bar opening prices, i.e. here it is not necessary for zero bar to be fully formed at all. Yes, testing would be really biased if one of the two remaining modes were chosen - since in the real world the signals in the zero bar would be constantly changing (perhaps there really would be a peek into the future here). It seems that the zero bar test can only be adequately performed in this test mode.

P.S. I agree with Yurixx's opinion. Rudeness should not be tolerated, although the expert should be acknowledged as very curious.
 
Yrixx, I completely agree with you about the lack of a common procedure (let's call it normalisation for clarity), for any existing turkey. Alas, I will say even more. Such a procedure, by definition, would be rather obfuscatory. And again, it would be entirely on the conscience of the author, as well as the idea of the indicator. What can you do, you know where the free cheese is... I will try to do it for my favorite Bollinger and Widget, and post the results here. Unfortunately, I have a lot of work to do here. I just crawled up to the monitor and I'm afraid I'll have to work at the weekend... Well, at least I'm glad the public didn't think it was a load of bullshit :)
123456789101112...28
New comment