From theory to practice - page 707

 
Novaja:

))

No kidding. Really, it's not clear what you mean when there is no single course for retail forex, there is no single course for real forex either? What are you asking about?
 
Vladimir:
No kidding. Really, it's not clear what you mean when there is no single course for retail forex, there is no single course for real forex either? What are you asking about?
and i don't get it ...
 
Maxim Dmitrievsky:

demolition is killed by differences, transition to increments, additive models. A few components are distinguished, conditionally stable (trend, seasonality, cycles). Next, we garchem. We just adjust models; if there is a pattern, it can be found, we don't even need to go into a terver...imho.

When using a ready-made model - you certainly do not need to go into, it is even harmful perhaps. But if you want to somehow change the model a little, it is hard to do without it. I want a model that harmoniously combines heteroscedasticity and persistence).

By the way, I wanted to ask you as an economist - what can I read of a general theoretical nature about central bank policy (creeping peg, etc.)?

 
Novaja:

Women's intuition has nothing to do with it. Signs of BP being "similar" to SB - there is a hospital average. The candlesticks are a quantization of the series in time with loss of information, no more, you can quantize in other ways. The distribution of the candlesticks? - It depends on the TF, the characteristics undergo changes, don't forget also that the information is somewhat distorted. However, the main trends of changes are traceable. Also series alignment by Erlang's flows is associated with partial loss of information, I came to it because the obtained series using this method is indistinguishable by its characteristics from the TF on larger periods. TF and ticks are different planes of one component. Fractality is inherent in random walks, it is inherent in BP too, only this principle is modified, but not lost. To give an example: a balloon, at first it is compressed, and when we start to inflate it, it expands, it is figurative.

women's intuition is exactly what it's all about..... women are able to listen to others first and then make a decision, unlike mature males who first prove they are right and then watch what happens ))))

The problem with all similar topics about "research" of time series is another one hundred millionth attempt to apply the school and university mathematical apparatus "to everything that moves, what does not move we push aside and investigate anyway". ))))

ZS: r/φ-features in Google search ;)

 
Igor Makanu:

The problem with all similar threads about "research" on time series is that this is yet another one hundred millionth attempt to apply school and university mathematical apparatus "to everything that moves, what does not move, we push it aside and investigate it anyway". ))))

Nothing is new under the moon. By the way, neither is your remark).

But to me it seems quite clear
That's all it's just a little bit of history repeating
.

 
Novaja:

Women's intuition has nothing to do with it. Signs of BP being "similar" to SB - there is a hospital average. The candlesticks are a quantization of the series over time with loss of information, no more, you can quantize in other ways. The distribution of the candlesticks? - It depends on the TF, the characteristics undergo changes, don't forget also that the information is somewhat distorted. However, the main trends of changes are traceable. Also series alignment by Erlang's flows is associated with partial loss of information, I came to it because the obtained series using this method is indistinguishable by its characteristics from the TF on larger periods. TF and ticks are different planes of one component. Fractality is inherent in random walks, it is inherent in BP too, only this principle is modified, but not lost. To give an example: a balloon, at first it is compressed, and when we start to inflate it, it expands, it is figurative


the "quantization" is the point, candles quantize (in this case "quantize" is a dirty word in relation to tick flow) ticks to such an extent that people get a desire to compare BP to SB... I can, as an example, quantise a sine wave in such a way that Euclid himself would not then recognise the resulting process graph.
 
Aleksey Nikolayev:

If you use a ready-made model, of course you don't have to go into it, it's probably even harmful. But if you want to change the model in some way, it is hard to do without it. I would like a model that combines heteroscedasticity and persistence in a harmonious way).

By the way, I wanted to ask you as an economist - what can I read of a general theoretical nature about central bank policy (creeping peg, etc.)?

no idea, what kind of economist am i - finance and credit, with examples of the soviet economy :) just a sympathiser

 
Martin Cheguevara:

There are a lot of talented people in this thread, so why don't we move on to practice at last?) And use the chance to use collective intelligence?

Because I don't think, I'm even 99% sure, that you'll find something dramatically better than this algorithm in the next 100 pages.

I could make it multi-timeframe, so that it shows all the latest levels of the last channel on all TFs.

But I already did it...it was of little use to the robot. That is why I simply threw the idea away and did not use it again.

There are 79 posts left (from page 687) until my prophecy: "In
 
Aleksey Nikolayev:

As I wrote above, it all comes down to a theorist, in this case a Nash equilibrium and some random fluctuations around it. The problem is that a purely speculative market and a market with a major player (such as a central bank) are very different games. The mechanism of transition from the former to the latter and vice versa is unclear.

You're looking in the wrong place, the TV corridor won't let you out...

Differential pursuit games. Here theorver may have an auxiliary value (in data preparation), although you can do without it -- everything you need can be provided by filters.

Problem statement :

.

Have you encountered such tasks?
 
Aleksey Nikolayev:

If you use a ready-made model, of course you don't have to go into it, it's probably even harmful. But if you want to change the model in some way, it is hard to do without it. I would like a model that combines heteroscedasticity and persistence in a harmonious way).


Ugh. I'm breaking a sweat.

Reason: