Machine learning in trading: theory, models, practice and algo-trading - page 2572
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
Thank you. I'll take a look, for starters. Maybe use it in some way I come up with
You're welcome, rather for what))) I spent a few days to find all of these links :)) Some are not displayed, the forum blocks them (
There is no point in looking, you need to parse and analyze how a number of
I think that in the beginning it is necessary to somehow calculate the statistics, whether it makes sense, and then implement in the training process.
So the question is still - how to do it right.
Suppose I have 3 such binary sequences with 10 measurement points at comparable time intervals.
A[]={1,0,0,1,1,1,1,1,0,1};
B[]={1,1,1,1,1,1,1,0,0,1};
C[]={1,1,0,1,1,1,0,1,0,1};
And so I want to understand/plot how the probability of a unit changing as the units increase in a row.
I understand that we need to count the number of sequences to begin with, but again, should we count the long sequences as one or should they be counted separately, for example 1111 split into 1,11, 111 and 1111 or is it just 11?
And then what to do - how to assess whether there is a regularity or randomness of the process?
To be honest, I don't understand much. The question is, does the probability change over time? To study this, you can simply construct a logistic regression on time (and check the significance of the difference between the coefficient and zero).
If other factors affecting the probability are studied besides time, you can also try to add them into a logistic regression.
You're welcome, or rather there is a reason))) I spent a few days to find all these links :))) Some are not displayed, the forum blocks them (
Seeing makes no sense, you need to parse and analyze how a number of
I looked at all of them. EurUsd most of them are now in the short, and two of them are in the long. Of course it is shorter. But as I understand it is by the number of traders.
For example
This info would be better expressed in lots. Because 1 trader with 100 lots may equal 100 hamsters with 1 lot.
Of course it is possible to parse, but there is no historical data.
Mikhail seems to have used the OM with CME in his article. And it seems that this OM can be found for many years. This is probably more promising, because you can collect information for many years at once. And it seems to be there in lots. I have to go read again.
And to collect on your own, you need a few months at least to have something to learn from.
Honestly, I don't understand much. The question is, does the probability change over time? To study this, you can simply build a logistic regression over time (and check the significance of the difference between the coefficient and zero).
Or maybe it's easier to make another predictor - the distance of the data string from the current one. Forest itself can calculate that data older than 8 months is bad for the current forecast. And there would be a simple split: before 8 months (with better leaves) and after 8 months with worse leaves.
Well on a tray they all learn well, of course. On the test/crossvalidation we need to check. But how? It's not clear. It is not even the importance of the predictor, but the importance of the split.
I looked at all of them. EurUsd is in the short most of them, and two of them are in the long. Of course it's shorter. But as I understand it is by the number of traders.
For example
This info would be better expressed in lots. Because 1 trader with 100 lots may equal 100 hamsters with 1 lot.
Of course it is possible to parse, but there is no historical data.
Mikhail seems to have used the OM with CME in his article. And it seems that this OM can be found for many years. This is probably more promising, because you can collect information for many years at once. And it seems to be there in lots. I have to go read again.
And to collect on your own - you need a few months at least, to have something to learn from.
That's right, do it, because you have to do what's easy, not what you have to do.
vladavd # :
elibrarius #:
I won't give the link.
))))
see my post on the previous page
Mikhail seems to have used the OM with SME in his article. And it seems that this OM can be found for many years. This is probably more promising, because you can collect information for many years at once. And it seems to be there in lots. I have to go read again.
And to collect on your own - you need a few months at least, to have something to learn from.
OI is interesting, but the problem is that the options thing is very complicated - to really understand net long vs. net short it is necessary to move mountains. Also, there are dark pools - OTC-clearing, where the trades sometimes make 30-50% of the volume (and maybe even more - who knows).
I'll tell you the main secret - practically all markets for retail move on the principle of reverse positioning of the majority. That's why this statistic will never be seen by retail. Order Flow sells for this reason as well
That's why retail will never see these statistics.
I do, even though I'm a retailer.
WallStreetTrader on pillion look it up, I'm not giving the link, they're going to rub the comment
What do they give? Don't they just copy the information from CME into their indicators?
As I understand it, their main trick is to identify large volumes that supposedly have been made by funds and other big players. I am curious, how do they distinguish them from the mass of other trades?
I found a video explaining how to use it. All they say is: maybe, most likely, they sold here because the price went down, etc. In my opinion - this is bullshit and we will see the same 50/50%. Search says that there was a signal of the same name here, but it is already closed, apparently merged.
And here in the main video from the supplier of the indicator is more beautifully explained, but apparently picked good moments on the chart.