a trading strategy based on Elliott Wave Theory - page 50

New comment
 
Pictures on this forum can be placed only one way:<br / translate="no"> [img]url of the picture[/img] . So you have to first place the picture on the Internet on any site, I have already shown a branch where you can do it, but it requires registration and confirmation (by a moderator). Or find any public resource where you can place the picture, copy the link to it and then again, place it in your post with the tags [img]. Any picture from Wordboard can be saved in "My Pictures" folder beforehand.


Ok, thanks for the help.

Regards,
Alexey
 
I'll try to post what I wanted tomorrow.
No time today...

Regards,
Alexey
 
Dear Alex Niroba, you'd better say something on the merits of the issue.

Just now Alex Niroba posted excerpts from the Diploma and now he has removed it again =) left one "-"

Dear Alex Niroba, keep to the point - share your strategy, what principles, what we use, how we use it, and you shouldn't rubbish with excerpts from the Diploma.
Sincerely, Davaeron.
 
Уважаемый Alex Niroba, Вы лучше что-нибудь по существу вопроса выскажите.

Alex Niroba just posted excerpts from the Diploma, and now he has removed it again =) left one "-"

Dear Alex Niroba, stick to the sub - share your strategy, what principles, what we use, how we use it, and excerpts from the Diploma not to flub.
Regards, Davaeron.




Davaeron,
The topic of my diploma is directly related to the issues discussed on this forum.
Today I do not have time to post pictures, and the text without pictures will not give a full
and vivid description of the issue at hand.
Therefore I will postpone the question until tomorrow.

Regards,
Sincerely Alexey.
 
<br/ translate="no"> But as it turned out, judging by the picture presented by Vlaidislav, in addition to the direct search for all channels in a row, he also specially selects those samples that start at the extremes of the channels. Hence, I think there is some discrepancy between what we get and what Vlaidislav gets. I am not sure yet that the selection of channels specifically by swings may give some additional forecast accuracy, but the search algorithm will become more complicated - this is for sure. So far I haven't got around to checking it. Although, logically, it certainly makes sense - all traders do so - they draw these swings on their terminals and look at them all the time. But of course, there may be some ambiguity about the largest channel. Only Vladislav can tell us if the largest channel is a real one or it is the best one that could be taken at that moment for the swing we have?


Sorry - I was at war with my EA again - it was getting pretty sluggish. So I had to go to "astral" for a while :). But now I've reduced the codes to 7.6K strings :).

Channels are high\low, intervals from 60 to 99.99%. There are no divergent channels in the figure. When there is a channel which is not defined as a convergent one, no matter how long it is, it is not taken into consideration when plotting the projection as channels tend to break through - i.e. the longer price has been in the channel, the more probable its breakthrough is. I do not have the exact criterion to determine that moment (I hope so far) - I use Murray levels, or wave-counting, or support/resistance levels, or something else I can think of.
2 Rosh Regarding the parabola - it has an unpleasant side effect - it is impossible to identify the extremum reliably, until it is actually passed. All because of such an unpleasant feature of the second-order line that this line can be drawn through two points in a non-unique way. So it is not very convenient for drawing projections. For confirmation of the passage of the pivot point, however, will do.
That's why I do not use the lines of the second order when drawing the projections.

Good luck and good trends.
 
I also agree with Yurixx, in the sense that approximation by a parabola is useless, just in my opinion Vladislav, using some way found quadratic function determines what is the potential energy in the channel, and thus selects channels with minimum energy (since any system tends to the state with minimum energy, and this channel will be the most stable and give the best forecast in the future).


Quite right - I wrote about it that a minimum of the potential energy functional serves as one of the criteria for channel selection. And this is a property of the potentiality of the price field, and I'm not looking for the trajectory itself due to (again) the fact that all trajectories that fit within the confidence interval should be considered equivalent for a given probability. That is, the construction of projections comes down first to sampling, then to linear algebra.

Good luck and hitchhiking trends.
 
IMHO, to assess the potential of a channel, you should also consider the time spent in that channel. The stability time of a channel depends on its angle of slope and its width (preferably via sigma). The steeper and narrower the channel is, the less time it will be stable. The time it is stable will be the CB, which can be estimated from the history. IMHO, the channel stability should be estimated at the moment of its intersection with the Murray level. The evaluation should not necessarily be continuous but discrete, for example: high, medium or low stability. Or is it all evaluated by the distance from the beginning of the channel formation to the Murray level?
 
All because of the unpleasant feature of the second-order line that it can be drawn through two points in a non-unique way.

You must mean 2 points in the literal sense of the word? If we take our case, then according to MNC for 1 price channel going from point A to point B only a single parabola can be drawn. That is, you need to change the length of the channel by at least 1 bar in order for the parabola to change its direction. Then, of course, this next bar could easily reverse the parabola.

I agree with the fact that linear regression is more convenient for prognosis, but for now I'm using parabola and linear regression channels for probability estimation too. Perhaps I will give it up in future, but I've decided to watch it too. Besides, if you look at a parabola having the length of at least one trading week, it does not change its direction so frequently. I do not consider shorter parabolas long time ago, because it is practically senseless due to the well-known property of sharp changes of direction on the next bar.

Although if your riddle about the mysterious quadratic function will be solved ;o), which everybody feels inside, but in reality nobody has written its formula yet and explained it here on the forum, then the available parabola will be some kind of auxiliary alternative to the quadratic function IMHO. Moreover it is unlikely that solving the final formula for the quadratic function and the method of its finding (optimization) can principally improve the results available with the parabola. Although I think its finding is already a sporting interest for all participants of this branch including me :o).
 
Here is the time trace from solving on each bar the problem of minimising the RMS of linear regression errors of samples from 60 to 1000 preceding bars. That is the history of finding the nearest channel.
The yellow line is the last linear regression value of the found channel, purple areas above and below are the 99th percentile of confidence interval.


Only I use the condition RMS 1/2 >= RMS 2/3 >= RMS to select converging channels.

The indicator in the bottom window is analogous to Rosh's indicator, but with the added RMS of 1/2. It is plotted for the last bar on the chart.

The open position is not mine ;-), it is ampir's.

As can be seen, the nearest channel may "twitch" considerably.
Does this picture look like the actual results? Maybe somebody has already got it? Maybe I'm dripping in the wrong place? ...

SZY: This picture counts quite slowly.
 
IMHO, to assess the potential of a channel, you should also consider the time spent in that channel. The stability time of a channel depends on its angle of slope and its width (preferably via sigma). The steeper and narrower the channel is, the less time it will be stable. The time it is stable will be the CB, which can be estimated from the history. IMHO, the channel stability should be estimated at the moment of its intersection with the Murray level. The evaluation should not necessarily be continuous but discrete, for example: high, medium or low stability. Or is it all evaluated by the distance from the start of the channel formation to Murray level?


The idea is certainly interesting and it occurred to me too. However, there is one subtlety. If I understood you correctly, CB is an average value. If it is (and even if it is not), I would like to know how you can use history to estimate channel stability ? The difficulty I see here is this.

As you rightly point out, there are two parameters on which the lifetime of a channel (most likely) depends - slope angle and width. If they didn't exist, you could just create a statistical series of all channels on the history and calculate the average and sko for it. And having a sko - estimate the probability that the lifetime of a given channel is out. :-) Then we could (just as we do now with Murray levels) draw vertical lines whose intersection with the channel lines would give additional information about the reversal zones for each confidence interval.

However, there are these two values - slope angle and width - which makes it impossible for us to compare the lifetime of the two channels if they have different values. I think a solution to this problem does exist, but a CORRECT SETTING OF THE PROBLEM is needed. As a person far from mathematical statistics I address to experts: dear Vladislav and others, maybe you will take the trouble to formulate the statement of this problem?
1...434445464748495051525354555657...309
New comment
Reason: