neural network and inputs - page 32
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
Why not? I wrote it all down at the top. Well, I've written it all down. There's a lot of confounding factors. It's not a sine... That's what I'm saying, "catching" Catching. Maybe it'll work. Maybe it won't. Everything in the market has a probabilistic nature.
You know, you're walking through a field. The wind is blowing, cars are driving by, birds are calling, there's a buzz. And then someone calls out to you. If you're near, you'll hear them. If it's far away, you won't. If it's in the middle, you'll hear it, but it's quieter... That sort of thing.
We really don't have any real volumes. So we have to calculate, on the strength of the reaction
Why not? I wrote it all down at the top. Well, I've written it all down. And there's a lot of confounding factors. It's not a sine... That's what I'm saying, "catching" Catching. Might work. Maybe it won't. Everything in the market has a probabilistic nature.
You know, you're walking through a field. The wind is blowing, cars are driving by, birds are calling, there's a buzz. And then someone calls out to you. If you're near, you'll hear them. If it's far away, you won't. If it's in the middle, you'll hear it, but it's quieter... (Like this.
)Maybe you will, but what? You analyse it for a long time and get a pattern, which is probabilistic. How do we determine that it is the result of action of a group of traders and not the manifestation of the cumulative effect of many factors? The positions of a hundred traders are opened within an hour, and each of them implements his or her own trading method. Randomly, the beginning of this movement coincided with the crossing of two wagons.
Another time, after crossing these two waves the total effect of opening or closing the positions of 200 traders who do not even think about these waves.
You will define these two movements as conscious trading of a group/group by a wave, when in fact it was an accident
So? Now imagine how many mashups there are - there will always be such coincidences and they will be random in nature and not the result of conscious trading of a particular group on a particular instrument.
come to the usual analysis of the TA tool
Yes, this is indeed the case... Only the key here is that we will apply TA to a BUZZ rather than to one or two MAs, and this bundle is monotonic and continuous (hypothetically)
)understandable
In short, if we assume that - there are groups, the curve is smooth, they trade continuously and round the clock, have an unlimited financial resource (on all instruments and on all TFs to trade), we know and know how, and now we simplify, and this neglect etc...... , in short, yes, yes, and in the end we come to the usual analysis of the TA instrument
There are no methods for calculating such groups without having a priori additional information
yes, there are, but not in forex - you need real volumes
+5
I would also add that if there is a group, it should be an alpha male in the market, i.e. able to move the market (like the turtles did in its time), and the orders of various competing groups will be spread almost equally on both sides of the price. Naturally at some moments the shares and stops will be distributed with some distortion that will cause a movement. Generally, there is information about the group actions in the price, but you're right. There are no methods for calculation of such groups without additional a priori information.
I understand that IronBird is your hypothesis for building your TS but you don't have a system based on this principle?
You and I have a misunderstanding because you think in terms of MT and I think in terms of Matlab. The difference is that in MT it's variables and in Matlab it's vectors. No offence, just trying to understand why things are so difficult...
Your last post is all wrong. There will be no testing there (again MT). There will be an algorithm that involves calculating the state of the MA bundle, and a solver that has to look for the dependence. All this stuff goes along the quotes chart and looks for regularities at EVERY BAR, or rather the weight (volume) of subgroups (instead of me looking for them long and hard in the toaster by the search method). And it either works, or it doesn't. If it works, it's not bad. If it does not work - then either the solver is wrong, or there are too small volumes on MAs (really, according to the market) - they are lost in the overall buzz (against the background of the volume of pipelines using all other logics). So either we continue trying to introduce something into the algorithm (for example, the logic of "defectors" from one subgroup to another), or we throw everything away, and think about how to divide the pipelines into other groups (rather than MA). And I wrote to you above that I use a breakdown into other groups altogether, and I took MA as an example.
------------------------------------
In short, if we assume that - there are groups, the curve is smooth, they trade continuously and round the clock, have an unlimited financial resource (on all instruments and on all TFs to trade), we know and know how, and now we simplify, and this neglect, etc...... , in short, yes, yes, and at the end we come to the usual analysis of the TA instrument
-------------------------------------------------------------------------
Of course the subgroups will not be static, the volumes in them will "float", but bet on the fact that it is in mass, as Vybegallo said, will not be too fast.
About market reaction - how to extract useful information from noise - I hope it is not necessary to tell here? I'm sorry, there is plenty of literature on the subject.
+5
I would also add that if there is a group it should be alpha male in the market, i.e. able to move the market (like the Turtles did in its time). In other words orders of different competing groups will be spread almost equally on both sides of the price. Naturally at some moments the shares and stops will be distributed with some distortion that will cause a movement. Generally, there is information about the group actions in the price, but you're right. There are no methods for calculation of such groups without additional a priori information.
I understand that IronBird is your hypothesis for building the TS but there is no system based on this principle.
The system itself does exist, only it does not catch mashechnikov. I took MA to show how to make an algorithm for determining this or that group of adherents of some paradigm (MA in this case).
You're talking about the alpha male there... Well, yes, that's true. But you should understand - here you are talking about quantitative, not qualitative. Well it will not work - it means that there are few of them. We are looking for a paradigm where there are many. And of course we will not find a breakdown that describes the entire market.
Don't worry - it all makes sense.
you analyze a bunch of mash-ups - everything is fine.
Why and for some reason you think you are calculating some mythical groups of traders, but you can think whatever you want - it's a free country
I don't analyse ma. I break them into groups because the curve is then smooth. It's them who break into groups, not me...
The system itself exists, but it doesn't catch the MAs. I took MA to show how to make an algorithm for identifying a particular group of adherents of a particular paradigm (MA in this case).
You're talking about the alpha male there... Well, yes, that's true. But you should understand - here you are talking about quantitative, not qualitative. Well it will not work - it means that there are few of us. We are looking for a paradigm where there are many. And of course we will not find a breakdown that describes the entire market.
The funny thing is, I showed him exactly how to catch the machers (see picture five pages earlier). In fact there is a digital FIR filter (neuron) from the signal from the two mash-ups. If you take at least a small fan of wipes and build similar neuron filters from each pair in the fan, then the total MO of such signal system (and all my signals are additive, i.e. summable, I do it that way as a matter of principle) will be much higher than the spread.