From theory to practice - page 274
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
At k--> to infinity we obtain analog of normal distribution, I propose to do otherwise, not to look for such k, but here and now to transform residuals which we have in a tail, we have not received them simply as a result of a delay on the way.
Perhaps you and I are talking about the same thing, only you from the non-entropy side and I from the Erlang side.
No, I don't get it.
I mean, we don't have a real quote for 20 seconds, for whatever reason.
I have about 10 pseudo-quotes with the same value written into the time series during that time. I support a Poisson flow.
How would you go about it?
As K increases --> to infinity, we will have fewer and fewer pseudo-quotes.
The Erlang flow will be of some K-order with an aftereffect, in which almost all quotes are real.
This is the kind of BP we can and should feed into neural networks - there should be awesome forecasts.
I do not see any other purpose of increasing K.
Ideally, from my point of view, K should be such, that ALL quotes in BP are real, but as little as possible, without increasing it further, so as not to lose accuracy of calculations
No, I don't understand.
I mean, we don't have a real quote for 20 seconds, for whatever reason.
I have about 10 pseudo-quotes with the same value written to the time series during that time. I support a Poisson flow.
How would you go about it?
Where will the intensity be greater? On the pseudo-quotes or on the real data?
Where will the intensity be greater? On pseudo-values or on real data?
I have rate (trading intensity) = T/N,
where T is the total number of ticks (real + pseudo), in fact it is the number of exponential time units,
N - real ticks.
This is a correction factor for the diffusion calculation.
If T=N, then correction factor = 1.
The higher the real values, the higher the rate of trade.
Ideally, from my point of view, K should be such, that ALL quotes in BP are real, but as little as possible, without increasing it further, so as not to lose accuracy of calculations
With greater k there is a problem with the timing of the decision to make a trade if the intervals are too large. Agreed.
My pace (trading intensity) = T/N,
where T is the total number of ticks (real + pseudo), in fact it is the number of exponential time units,
N - real ticks.
This is a correction factor for the diffusion calculation.
If T=N, then correction factor = 1.
The greater the number of real values, the higher the rate of trade.
Do you add the pseudo number to the number of real ticks and divide it all again by the number of real ticks? Could there be some kind of mistake or a typo? Then it turns out that the more pseudo values, the higher the rate of trade.
Do you add the pseudo number to the number of real ticks and divide it all by the number of real ticks again? Could there be a mistake or a clerical error? Then it turns out that the more pseudo values, the higher the tempo of trades.
Nah, I'm not adding anything. I set the size of sliding observation window = 10.000 (for example). In it I count the number of real (say = 5000) and pseudo (say = 5000) ticks. I get the trading rate = 2. I.e. 1 real quote per 2 time units. The rate is the inverse of the speed.
No, I'm not adding anything. I set the size of the sliding observation window = 10.000 (for example). In it I count the number of real (let's say = 5000) and pseudo (let's say = 5000)- ticks. I get the trading rate = 2. I.e. 1 real quote per 2 time units. Rate is the inverse value of speed.
In this case, how is the situation of delayed ticks receipt handled, but in a batch and such incident appears in the observation window? recalculation based on ticks time?
how is the situation of ticks arriving with a delay, but in a batch, and such an incident is included in the observation window handled? recalculation based on tick times? or how
Suppose the exponential number generator generates 5.
After 5 seconds, the last incoming quote is read. If its time differs from the previous one, it is a real quote, if not - a pseudo-quote. There can be any number of ticks in 5 seconds interval, we do not care about it.