Machine learning in trading: theory, models, practice and algo-trading - page 1026
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
)))) omg...
Why didn't you buy it? are you stupider than a horse?
what is predictable
then why not)
What makes you think I didn't buy it?
What's so tricky about it? The bounce from the lower limit of the channel, it was clear to the horse).
Post factum you can draw anything, that's what technical analysis is all about.
Everything works, trends, flutes, etc. ))))
Do you know what these prices are?
here is a simple code
price <- cumsum(rnorm(500))
plot(price,t="l")
If you don't understand it's a randomly generated quote
People often see what they don't see, they see what they want to see.
What makes you think I didn't buy?
It's easy to check) just show the deal)
Post factum can paint anything, that's what technical analysis is
.....In case you don't understand, these are randomly generated quotes.
They are far from random.
kotir goes from and to and does his puppet thing
My semi-automatic machine works on crosses, because it uses averaging, I do not trade majors. I set the trend direction, the rest is done by the machine. This is the time of the deal, when the EUR turned up on Thursday on EURAUD cross:
Thank you! Until now I did not believe that you can make money on classic TA, you are just a hero!
they are far from random!
kotir goes from and to, probably on a random path
Renat, not the quotes are random, what is in the picture is random
Renat, it's not the quotes that are random, it's what's in the picture that's random
aaaa,
fact
----
And how does a level appear in such chaos, how do you train a machine?
According to your logic, I need to put all sorts of theanalysis into the algorithm and do the opposite?
-----
Well, to be sure, I will show mine, so you do not think, as usual in this branch zasazhivayut my questions, they are not for themselves, but for the sport of interest:
Look who's talking at once!!!!! The main thing is to get started.
Indeed, the nature of the market is such that its understanding can go against logic and common sense. Why? Because there is no such thing, no nature, law or pattern. The market is what it is here and now and why is it so? The answer to this question is very simple. Because the balance of opinion between buyers and sellers at this moment in time was like this, this is how it looked during this hour, this is how it looked during the next. Well it's all fantasy.... Philosophy!!!!
Post factum you can paint anything, that's what technical analysis is
everything works, trends, flutes, etc. ))))
Do you know what these prices are?
here is a simple code
price <- cumsum(rnorm(500))
plot(price,t="l")
If you don't understand it's a randomly generated quote
People often see what they don't see, they see what they want to see.
Buddy, excuse me, but I'll answer you with a quote from the famous cartoon "You're a jerk!!!!" (Why? Because your example shows that you don't know how to use this analysis. As shown in your screenshot, BUT finished with red lines, I personally see potential targets. And if you draw it as you did, there is really nothing there. But as you can see, the technical analysis can be not only post factum.
The problem of technical analysis is that it always considers two scenarios. And as a rule the market requires the unambiguousness of the answer. That is why many of the inhabitants of this area use the AI system.
That's another thing you mentioned about levels. I tried to cram the values of levels into NS a long time ago, but nothing useful came out. But that was then. Now the level of knowledge is completely different and I think that the subject is very perspective. But first let's formulate the problem. Who first??????
Who's first??????
We should get a bunch of indicators that form levels, and try to determine from which level of the set there will be a bounce. So, the classification should work.
It is clear that if you have proposed to me to start.
We want the network to give us the value of the level - BEFORE the number, let's bother with normalization and subsequent reverse interpretation.
We build the level independently and the network determines its significance - quite a suitable task, but for this you need to have an algorithm for building levels (basic strategy). Manual construction of levels is excluded, since the collection of statistical data for training can be in error, which will eventually affect the result of training. To determine the significance you need a complex network structure for ranking. The problem is interesting but as a subtask of the main other, more important problem..... namely.
Having received a level from the underlying strategy we determine its status as to whether it is a support or resistance level? Yes, no? - this is the fundamental task of working with the levels. Having received the level, we want to know if it is support or resistance and it does not matter where this level is above or below the price. After all, in the market it is considered that if the level is higher, it is resistance; if it is lower, it is support. Here we break this stereotype with the help of AI. Having received the level below the price and defined it as resistance, rather than support, we will be at a better price than the price level that can be taken as a signal for the continuation of the trend. I will give you an example and for that I will dig up my ancient but quite obvious candlestick indicator that is able to form levels.
The candlestick pattern forms a level. If the level is broken through, the line becomes dashed, as with the "shooting star" example, the red ones form resistance and the blue ones form support. Now imagine this situation.
at time where the number one is, we identified this line as support and colored it blue, then it turns out that all the time we were in the area of better prices than the level and any purchase at this level would be profitable, because by identifying the level as support we would potentially see the growth of the rate as a prospect and in this case would be right (the forecast worked) But everything is not so simple alas ......