Comparison of two quotation charts with non-linear distortions on the X-axis - page 8

New comment
 
Integer: Got something... just didn't realize it. I'm experiencing a sense of mystical ecstasy :)

^_^, thank you, that's at least some result, because I've been googling for two nights now - it won't admit it, it keeps going in the direction of neural networks....

..... in 4 hours.

I was surprised to discover that Delphi already has Bezier methods in Canvas property, which calls gdi32.dll, my first experience with Bezier plotting: http://imglink.ru/pictures/10-03-12/aa4a960cdbb0cc6709a480432f9772b1.jpg, it reminds me of something :)

 

Получилось кое-что... только не осознал. Испытываю чувство мистического экстаза:)

Very interesting indeed. The topic is very interesting. And the interest is both academic (dispelling the pattern theory) and purely practical. Can you describe in more detail what I got out of it?

my first experience with Bezier: http://imglink.ru/pictures/10-03-12/aa4a960cdbb0cc6709a480432f9772b1.jpg, it reminds me of something :)

We understand, of course, it's March 8)))

 
anonymous:

2. approximate each of the bars by a polynomial of third degree (for bar C>O by points: {(t[0];Open), (t[1];Low), (t[2];High), (t[3];Close)}, for bar O>C by points {(t[0];Open), (t[1];High), (t[2];Low), (t[3];Close)}, where t[i]=i/3)

I've been thinking about your suggested variant, but now I'm inclined to think that it is fundamentally wrong to "mix the OHLC data" together

For the time being I'm inclined to think that man when searching for patterns, pay attention to contours that are formed from above and the pattern base (figure) is Low of the first and the last bar, i.e. he sees a certain integral of a complex function in its physical representation - the area

ZSY: a lot of information about vector graphics, an interesting format SVG, all information encoded in text form, area seems to clearly define the figure, who found the source code to work with SVG ?

оригинал: http://imglink.ru/pictures/10-03-12/118db66e279940cda9327b8a516be82b.jpg

SVG : http://imglink.ru/pictures/10-03-12/2c9af48a73921a361d75dfec31f30795.jpg

Files:
lgtvlm.zip  10 kb
 
sayfuji:

Can you elaborate on the output?

The output is the yellow line, which is an orange line close to red.
 
What does it show? How does it count, by DTW or so? The top left corner doesn't change anything for some reason.
 
sayfuji:
What does it show? How does it count, by DTW or so? The top left corner doesn't change anything for some reason.

This is a test of the DTW function. The value at the start and end of the yellow line will always be as at the start and end of the raw data (orange line).
 
Sorry about the type of dialogue, and I've been following the thread, but I'm asking as a lamer, don't mind me. Which part is responsible for the prediction? The yellow thin one?
 
There is no prediction, it's just a function of stretching some data over others, so that they are as close to each other as possible. It's just a function, maybe someday we'll add it to something. I personally don't have such a task yet, it was interesting to figure out the algorithm.
 
That's what I thought, I just thought I was dumb. For those of you who are good at mathlab, here's a good link - of course you can see everything graphically.
 

I can't say how much pattern trading is effective for MTS, I started to create a system for recognizing patterns of, say, market behavior, in the optimizer got good parameters for one month on the H1 TF:

open_pattern=7 close_pattern=18 TakeProfit=980 StopLoss=490

http://imglink.ru/pictures/22-03-12/1f1f3e06c9143190e5e943ab51abff45.jpg

month forward: http://imglink.ru/pictures/22-03-12/366498d527e9000c9adb6814207ba6fe.jpg

10 trades per month http://imglink.ru/pictures/22-03-12/55c3ac7d1901acf0f0b4e6a0d0b9909f.jpg

123456789101112
New comment
Reason: