What is everyone looking for? - page 16
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
We all have one object - BP, but nobody asks the question about the number of orthonormal parameters to identify this BP. Actually it is not necessary. We need an orthonormal set of indicators for the TS which would have a forecasting property for the sufficient number of bars for profit withdrawal. But the idea of orthonormality is just a common culture of selection or elaboration of indicators. Everyone who has worked in Metastock has this level, but with metaquotes it does not. This is not the first time I've written about this. I don't care, though I've already implemented the algorithm. In MQL5 they use some programmer's rubbish instead of orthonormality and are happy about it.Please give me the definition of orthonormal indicators or parameters
Well BP has VERY few independent properties. The first derivative, and the spectrum, well the spectrum itself includes the amplitude at a certain frequency - so we have essentially mashups (spectrum properties) and the first derivative. And that's it. The first derivative is calculated as (High(i) + Low(i))/2) - ((High(i-1) + Low(i-1))/2). Well, (Open-+-Close) is in fact the most probable value of BP. And I wouldn't take Open and Close into account - it's in sperktra.
In Fourier spectral analysis, yes. It is argued that frequencies need to be bound to time, that a change in trend is preceded by a change in phase and something happens with autoregression etc. At a glance, spectral analysis is a part of it, and where is the volume for example. But at any rate - it's a conversation piece.
In Fourier spectral analysis, yes. It is argued that frequencies need to be bound to time, that a change in trend is preceded by a change in phase and something happens with autoregression etc. At a glance, spectral analysis is a part of it, and where is the volume for example. But anyway - it's already a conversation.Yes - also volumes. But there is a question of reliability. Alas.
Please give a definition of orthonormal indicators, or parameters
You cannot derive another indicator from one indicator in an analytical way. You cannot get volume indicators from a trend indicator and you cannot get volatility indicators from both of them. Hotf everything can be obtained from the original BP.
В Метастоке все индикаторы разделены на группы: трендовые, валонтильности, момента, объема, перекупленности-перепроданности и еще что-то - всего шесть групп. Подозрения, что это независимые индикаторы. В этом же Метастоке утверждается, что хорошая ТС должна содержать индикаторы из каждой группы. Когда-то я уже возмущался, что набор индикатор метаквотов - это от балды, мысли никакой. Кто-то из авторов любит и все. В результате все население этого кворума не знакомо с идеей ортогонального набора индикаторов.
Yes, there have been talks like that. And I think I even spoke out.
The point is that a person who knows the basics of TA and has an idea of what indicators show and how they are counted, does not need this division. But it will not help to "young Michurinians" who like to make it by gut feeling.
Otherwise, yes, such a division would look logical. But... it's not a matter of principle - it's not worth worrying about.
===
Just for the record: I am a long-time user of Metas, but, you know, I never had the feeling of deep satisfaction with segmentation, nor have I been modelling TCs from different indicators, like pies. Again, you seem to be exaggerating the problem.
Well BP has VERY few independent properties. The first derivative, and the spectrum, well the spectrum itself includes the amplitude at a certain frequency - so we have essentially mashups (spectrum properties) and the first derivative. And that's it. The first derivative is calculated as (High(i) + Low(i))/2) - ((High(i-1) + Low(i-1))/2). Well, (Open-+-Close) is in fact the most probable value of BP. And I wouldn't take Open and Close into account - it's in sperktra.
One indicator cannot be used analytically to produce another indicator. You cannot get volume indicators from a trend indicator and you cannot get volatility indicators from both of them. Hotf everything can be obtained from the original BP.
what you have described are not properties of a price series, but properties of a time-quantized price series. Quantization is essentially an indicator in itself. There are only 3 fixed-period quantizations in price, time and volume. But there can be a huge number of further transformations of the obtained series. Yes, most of them will be strongly correlated, because their input is the same, and there are not so many arithmetic and logical operations. Although, by sequentially applying them we can obtain any number of indicators. But they will be useless not due to the fact that they show the same things, but because they show nothing that is inherent to the underlying real factors of pricing.It's not quantization, but bar characteristics. It's true that there are more of them, but the essence is different. An orthonormal set of indicators is an element of culture, but there is also an element of education. There are different mathematical approaches to BP. The best known and most common is Fourier. A completely different approach is NS. What is needed? That's a very valid question of the topic.
Yes, there have been talks like that. And I think I even spoke out.
The point is that a person who knows the basics of TA and has an idea of what indicators show and how they are counted, does not need this division. But for "young Michurinians" who like to make it by groping, it will not help.
Otherwise, yes, such a division would look logical. But... it's not a matter of principle - it's not worth worrying about.
===
Just for the record: I am a long-time user of Metas, but, you know, I never had the feeling of deep satisfaction with segmentation, nor have I been modelling TCs from different indicators, like pies. Again: you seem to be exaggerating the problem.
In particular, with my participation in the forum I'm trying to raise its overall level, at least as far as I'm concerned. Everything to do with the terminal is a lower bar of qualification. By the way, if it's about the indicator set. I've never tried it and I've never read the documentation, it says the indicator's source code is in kodobase. That's a lie. There is no relation. One name, but the texts are different. Further. If the MACD is available, the authors must guarantee that Appel is the author and not a local genius. For all indicators there is no such assurance. Professionals do not allow themselves such blunders.
Rather than quantisation, it is the characteristics of the bar. It is true that there are more of them, but that's not the point. An orthonormal set of indicators is an element of culture, but there is also an element of education. There are different mathematical approaches to BP. The best known and most common is Fourier. A completely different approach is NS. What is needed? That's a very valid question of the topic.
In general, what people need is a nice arrow indicator that will constantly give sell signals near future haves and buy signals near lows :)