Please tell us your opinion. - page 4

 

And I did. That's why I stated all of the above.

At 20-50 trades in the optimization (I've tested it using a primitive Expert Advisor based on stochastic signals, 5 parameters).

further (outside the sample) there is an unambiguous loss.

Let's take a 2-3 times larger history. Optimize.

At 100-150 trades (on the optimization area), - further, outside the optimization period, - the next dozen or so trades more often (60-75% of cases) give profits, than losses.

And if deals of 300-600 at optimization period, then two/three dozen of following deals (out of sample) give profit in total...

Of course, this is my personal observation - based on my own several experts. But in general, I think the trend is true for most designs...

//----------------------------------------------------------------------------------------------

I should add that when I added the option to close positions by indicator signals, the profit and profitability increased a little! But it is important to choose the right closing indicator to fit the strategy algorithm.

 
rid:

And I did. That's why I stated all of the above.

At 20-50 trades in the optimization (I've tested it using a primitive Expert Advisor based on stochastic signals, 5 parameters).

further (out of sample) there is an unambiguous loss.

At 100-150 trades (in the optimization range), further outside the optimization period, the next ten trades (60-75% of cases) return profit rather than loss.

And if deals of 300-600 at optimization period, then two/three dozen of following deals (out of sample) give profit in total...

Of course, this is my personal observation - based on my own several experts. But in general, I think the trend is true for the majority of constructions...

Yes, the thought is interesting. Thanks for the replies.....)))))

 
rid:

And I did. That's why I stated all of the above.

At 20-50 trades in the optimization (I've tested it using a primitive Expert Advisor based on stochastic signals, 5 parameters).

further (outside the sample) we have an unambiguous failure.

Let's take a 2-3 times larger history. Optimize.

At 100-150 trades (on the optimization area), - further, outside the optimization period, - the next dozen or so trades more often (60-75% of cases) give profits, than losses.

And if deals of 300-600 at optimization period, then two/three dozen of following deals (out of sample) give profit in total...

Of course, this is my personal observation - based on my own several experts. But in general, I think the trend is true for most designs...

If we take into account profit and number of trades, then maybe the profit/number of trades ratio is relevant? Have you tried dividing the profit/number of trades?

 

No, I haven't. I can't get my hands on everything... (Again, domestic matters distract me - there's no getting away from them either!)

Besides, I (again) experimented with my Expert Advisors, where the leading source of signals for market entry was a stochastic indicator. For each trading tactic, the results are likely to be very different, even if the general trend is the same.

One more thing. I initially do not look far into the future in the area outside the optimization period, i.e. outside the sample. I am satisfied if several dozens of trades out of the sample yield a profit in total. After that the system may be optimized again. And so on...

 
rid: ..For each trading tactic, the results are likely to be very different...

Confirmed, they are different. Had this tempting idea myself, but the inspection did not confirm it. Even the WPR expert,

practically a stochastic, gives different results just after the end of the optimisation sample, if at that moment

... the market character changes (trend-flat, etc.).

 
LeoV писал (а): Have you tried dividing the profit/number of trades?

As far as I understand, this value is what is reported as the expectation of the deal. Or have I, as usual, got in the wrong way with my flubbing?

 
Mathemat:

As far as I understand, this value is what is reported as the expectation of the deal. Or have I, as usual, been out of line with my flubbing?

I did not know. Expectation = profit/number of deals?

 

The result is nice, but only the market knows what will happen next. For two years I have been writing EAs based on the same principle and I have noticed one peculiarity - abrupt alternation of profitable and failing sections. Sometimes it seems to be a real grail, half of a year of forward deals is so beautiful, and the next six months... My range with 50 profitable trades may be changed by another one with losing trades. Why is it so? I don't understand it completely, but I suspect that from time to time the market acquires a kind of "spontaneous nature", when its behavior slightly depends on the previous period, which is used by NS for learning/learning/adaptation (depending on the realization).


In any case the result looks good, and chances of robustness are there, don't listen to anyone :) (IMHO).


Just in case, is this a development in MQL or, knowing your interest in NSDT, using it?

 
Figar0:

Just in case, is this development in MQL or, knowing your interest in NSDT, using NSDT?

Of course using NSDT. All the optimization is done in NSDT. Although MT4 is better at picking losses and profits.....

 
Mathemat:

As far as I understand, this value is what is reported as the expectation of the deal. Or have I, as usual, got off on the wrong foot with my flubbing?

Nah, that's not what expectation is.....

Expectationof winning is the mathematical expectation of winning. It is a statistically calculated figure that reflects average profit/loss per trade. It can also be thought of as reflecting the expected profitability/loss of the next trade.