Machine learning in trading: theory, models, practice and algo-trading - page 3304
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
Back then, brokers did not collect tick values yet. I did it myself. I collected real ticks and stored them in files in portions for about 6 months. I applied them on the tester and got a completely different picture.
You are lucky that you decided to do a sensible business once...now brokers do not collect tick quotes anymore, but take/supply them from somewhere; it may be connected with MT server updates.
Sometimes you can see with the naked eye - here are fresh ticks, and here is a general bullshit. And even with minutes
in fact, the archives of real ticks as they really were, came from a particular DC - a very expensive commodity. (even through the underlining)
I'm surprised you're surprised.
you are lucky that you decided to do a sensible business once...now brokers do not collect tick quotes anymore, but take/supply from somewhere else; it may be connected with MT server updates.
Sometimes you can see with the naked eye - here are fresh ticks, and here is a general bullshit. And even with minutes
in fact, the archives of real ticks as they really were, came from a particular DC - a very expensive commodity. (even through the underlining)
You are wrong. Each broker itself collects real ticks and on the next day's tester you can get the real ticks of the previous day.
And exactly those ticks, which after filtering are submitted by the broker to MT5. You can check it in the "Every tick based on real ticks" mode, or copy it.
A good paper on how to properly do BP representation for neural networks. FFTs can be got rid of, of course. And further comparison of different models.
The fundamental difference is that preprocessing is built into the network architecture. But it can be done separately.
LSTM smokes in the background, because it does not take into account interperiod variations.
Bousting is also somewhere near the bottom of the ranking, based on their tests.A good paper on how to properly do BP representation for neural networks. FFTs can be got rid of, of course. And further comparison of different models.
The fundamental difference is that preprocessing is built into the network architecture. But it can be done separately.
LSTM smokes in the background, because it does not take into account interperiod variations.
Bousting is somewhere near the bottom of the rankings too, based on their tests.No, I understand that the more you search (iterations) the more likely you are to find something random that looks like something that is NOT random.....
But if we came up with an idea, and then matched the parameters to it in 10 iterations instead of 10000, can it be considered an untrained model?
After all, the very phrase"we came up with" also implies some kind of thought process (iterations).
How does the final model know whether it was brain or computer iterations and whether there is a difference between the two?
The question arose after reading Prado's article
Not so long ago on the forum someone gave the name of the effect (I haven't found it yet), because of which series close to SB seem to have a period. This effect is associated with a lot of shameful moments in science, when by means of Fourier "found" periodicity in processes, and radio amateurs because of it on the forum will never survive).
How to prove the opposite?
In my opinion, there are events tied to time - the same news. I think if we divide them into three sub-samples - expected, worse, better and take into account the context, we will notice similar behaviour of market participants.
Another option is seasonality of goods.
After all, the very phrase"we came up with" also implies some kind of thought process (iterations).
How does the final model know whether it was brain or computer iterations and whether there is a difference between the two?
The question arose after reading Prado's article
Overlearning arises from the memorisation of rare phenomena. These phenomena are isolated purely statistically, as there is no model describing cause and effect.
That said, a loss does not always mean that the model is overtrained.
Not so long ago on the forum someone gave the name of the effect (I haven't found it yet), because of which series close to SB seem to have a period. This effect is associated with a lot of shameful moments in science, when by means of Fourier "found" periodicity in processes, and radio amateurs because of it on the forum will never outlive).
Scale invariance?
Moire effect :)
Scale invariance, it seems, can be described in such a way, and will not be the subject of radio amateurs' research. Only without Fourier, but market periods like hourly and daily, which describe different activity.