Machine learning in trading: theory, models, practice and algo-trading - page 836
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
For the trader it's a competition, which teaches its models and gives them, lucky - good, unlucky - nothing to lose.
You have the wrong associations. If your model wasn't accepted because of overfits or undertraining, it means that you have to improve your skill. Luck has nothing to do with it.
You have the wrong associations. If your model isn't accepted because you're overqualified or undertrained, it means you need to improve your skill. Luck has nothing to do with it.
How long do models last? If they have competitions all the time, is it short term?
And how much money does the fund make in %?
A new tour each week. In a week you have to train the model and send them the predictions. But the forward estimate of your model will only be known after another three weeks, your predictions will be compared to the real ones for those 3 weeks.
I think they're keeping at least 90%.
:))) I will start to reread your topic after I finish studying RL
And if we assume that your predictors will be better than mine, it will be fine
Maxim, the attached file contains the BP for AUDCAD obtained at exponential intervals of ticks reading (to be more exact - the discrete geometrical distribution at n=0.5).
Column A - Bid
Column B - Ask
Column C - Intensity in the sliding window = 10.000
Column E is the timestamp.
Where the timestamp =0 means it is an artificial pseudotick.
I.e. there is still a real BP "sitting" inside this pseudo-random.
Can you extract real BP from source BP and feed it into the neural network with 2 BP returnees? One - original (pseudo+real), the second - only real.
Interesting.
When working with initial BP (pseudo+real), you have to realize that you are working with the simplest thread with no memory
Step 2. In this initial BP, you should take only every 2nd quote. You'll get a 2nd order Erlang flow with consequence. Check.
Step 3. In this initial BP you should take only every 3rd quote. You will get the 3rd order Erlang flow with consequence. Check.
Etc.
If you get something unbelievable, you owe me a signal.
Maxim, the attached file contains the BP for AUDCAD obtained at exponential tick reading intervals (more precisely, the discrete geometric distribution at n=0.5).
Column A - Bid
Column B - Ask
Column C - Intensity in the sliding window = 10.000
Column E is the timestamp.
Where the timestamp =0 means it is an artificial pseudotick.
I.e. there is still a real BP "sitting" inside this pseudo-random.
You can extract real BP from source BP and feed it into the neural network with 2 BP returnees. One - original (pseudo+real), the second - only real.
Interesting.
I'll try to shove it into ns tomorrow)
I'll try to put it into the ns tomorrow.)
I've added the check algorithm there as well. Be careful, please.
I also wrote the check algorithm there. Be careful, please.
Yes, I see that it will be more complicated with ticks, but I'll do it carefully.)
Convert these sequences into custom symbols of MT5, thus creating individual ready-made symbols... if they work
Yes, I see, it will be more complicated with ticks, but I'll do something neat.)
These series should be converted into custom symbols of MT5, and then you'll get some ready symbols... if they work
Well, you can do the opposite - first select the Erlang flow of order 100 and move down to the simplest :))
On the subject of predicting volatility. Let's say, predicting volatility is much easier than the quote itself
And there are even all sorts of models like https://en.wikipedia.org/wiki/Markov_switching_multifractal
What does it give, how to use it correctly, has anyone ever done it?
On the subject of predicting volatility. Let's say, predicting volatility is much easier than the quote itself
And there are even all sorts of models like https://en.wikipedia.org/wiki/Markov_switching_multifractal
What does it give, how to use it correctly?
GARCH is called, as opposed to machine learning, mainstream at financial markets (along with cointegration and portfolios).
The models take into account a bunch of statistical nuances of increments, including thick tails and long memory a la Hurst.
For example, there is a publication about choosing parameters of GARCH models on ALL stocks included in the S&P500 index!
There are a lot of publications on application in Forex. A very well developed toolkit. For example, the rugarch package.
So, let's leave the farm, go to the highway and go to the march "Farewell to Slavyanka".