The essence of optimisation - page 5

 
toxic:

Yes indeed, let it not be MO but Sharp, or MO/SCO.

I wouldn 't risk betting such a strategy on the real. Such noisiness is a sign of a shitty forecast or extreme MM like martin, because it looks like SB. Although...

I've posted a picture of tests without martin, it looks like the noise.

https://www.mql5.com/ru/forum/6919/page42#comment_540465

So what's the catch? You pick a maximum and trade.

Так ли плох мартин? Или нужно уметь его готовить?
Так ли плох мартин? Или нужно уметь его готовить?
  • www.mql5.com
, ком как), кривая балланса приобрела чуть более красивый вид, да и отдельные показатели тестирования слегка улучшились.
 
anonymous:
If it is Sharpe ratio - there is no point in selecting any point at all, as there is a dominant portfolio.

So the dominant portfolio is the one with maximum Sharpe, as far as I know, or not?

Or do you mean that you have other strategies with better characteristics?

In this case it's a mental experiment. Let's assume this is the only strategy.

nowi:
And what conclusion have you come to?

It's too early to talk about it, it needs to be approached, too new things cause a negative reaction as a rule.

Alex_Bondar:

I wouldn 't risk putting such a strategy on real. It's too noisy, it's a sign of a shitty forecast or extreme MM, like martin, because it looks like SB. Although...

I've posted a picture of tests without martin, it looks like the noise.

https://www.mql5.com/ru/forum/6919/page42#comment_540465

So what's the catch? You pick a maximum and trade.

I see.

No catch so far, I just want to understand the situation and have a betteridea of what I am trying to achieve.

 
toxic:

So the dominant portfolio, is that with maximum Sharpe, as far as I know, or not?

Yes, with a maximum. But that maximum will be greater than any of the highs from your chart.

It's too early to talk about it, it needs to be approached, too new things cause a negative reaction as a rule.

A negative reaction to new things is scepticism triggered by a sense of common sense. Spit it out, we're ready for your 'new' and holding on to our chairs tightly :)

 
toxic:

I see.

No catch for now, I just want to understand who's view of the situation so that I can make my idea clearer.

Does this curve even make sense? You've been quick to retreat from MO to Sharp ratio.

Take a normal (real) BP, strategy, test, get ratios, then we'll talk, I can not think about "mental experiments" and left-handed curves, I'm through the roof.

So tell me what you have there. After such tedious prelude, I don't count on less than a FREE grail.

 
toxic:


 
pagot:
Not for all, but for most scalpers optimization makes no sense, because the results in the tester and on the demo (real) are very, very different.
Besides, scalper results in the tester depend on test modes: normal mode, with random delay, all ticks, OHLC on M1.

I agree... As a rule scalp makes no sense at all in long term mode

But if a scalp strategy survives with 4-5 times the spread, that's super

that's why it makes sense to test it under tough conditions.

 

I would do the following:

Apply Gaussian smoothing and take all obvious extrema >0.5 on them to make a portfolio (weighting the lot according to their value(profit/risk).

There is no special point in selecting the maximum itself, due to noisiness.

This is what I meant by "smooth hyperplane" as a one-dimensional case.



 
Optimisation is about getting information about some lucky points of contact between your code and one of the manifestations of reality in the past. Expecting these fortunes to be repeated in the future is like looking at a 5-year-old child's autumn landscape drawing in winter to find the area in which it was made.
 

Since no one believes in optimisation, why should I throw my weight around in front of you gentlemen?

I don't see the point, not yet.

None of you are not even close to my understanding, I'm sorry, but I cannot tell everything from scratch, if only there was some approximation, a foundation of understanding, it was still possible, but alas for now.

I will try in about five years.

 
toxic:

Since no one believes in optimization, why should I throw beads here in front of you gentlemen ...?

It's just a fashion. It's a new one. To shit on those who have a lower rating. Do not worry, save your rating. Or come back in five years, maybe the trends will change)).

No kidding.