Machine learning in trading: theory, models, practice and algo-trading - page 1169
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
through the model we go through the increments
Exactly like this and in no other way.
There must be precise criteria for excluding a particular increment from the set, or even a particular sample of increments from the analysis.
I use the kurtosis and skewness of the distribution of the moving sample increments for this purpose. If the current sample does not meet some table values, I do not consider it at all.
I consider this one of the working methods of data preprocessing - there may be others.
The results of my TS have improved tremendously - but, I'm not happy yet, I've been burned so many times... I am waiting for the week to come.
In short.
Once again I say to the most stubborn neural networkers.
Whether you like it or not - a neural network, like any other type of BP analysis and prediction, can work exclusively and only on stationary series. Do you, with the lowest education, think you are smarter than Kolmogorov!
So be kind enough to do pre-processing of BP - exclude from it those increments that lead to non-stationarity of the process - namely, anomalous outliers in the heavy tails of the distributions.
If you are sure that any kind of smoothing/filtering leads to lagging - just exclude the anomalous data or data chains from the BP and that's it.
Amen.
It's not really about practice. Such charts create some connection between the market and the "coin exercise".
Piecewise-stationary SB does not simulate a flat well. An obvious development of the model is the piecewise homogeneous Markov chain. It is quite possible to use MO to train it. Surely someone has done something similar, but I haven't found anything.
In short.
Once again I say to the most stubborn neural networkers.
Whether you like it or not - a neural network, like any other type of BP analysis and prediction, can work exclusively and only on stationary series. Do you, with the lowest education, think you are smarter than Kolmogorov!
So be kind enough to do pre-processing of BP - exclude from it those increments that lead to non-stationarity of the process - namely, anomalous outliers in the heavy tails of the distributions.
If you are sure that any kind of smoothing/filtering leads to lagging - just exclude the anomalous data or data chains from the BP and that's it.
Amen.
Neither anomalous outliers nor heavy tails mean that the time series is non-stationary.
Outliers are generally a derivative of a particular predictive model, and tails don't matter if the distribution function is constant.
Don't bullshit.
Neither anomalous outliers nor heavy tails imply non-stationarity of the time series.
Outliers are generally a derivative of a particular predictive model, and tails don't matter if the distribution function is constant.
Cut the crap.
Whatever you say, pops. Whatever you say... I bow my head to your genius.
In general, I would recommend that dads and the rest of the old people who populate the forum attach the state to their posts.
So everyone can see the IQ of this elderly child.
This forum will have some much-needed fun and banter.
In general, I would recommend that dads and the rest of the old people who populate the forum attach the state to their posts.
So everyone can see the IQ of this elderly child.
This forum will have some much-needed fun and banter.
If you're talking about me, then my real account post was published many years ago on foursquare.
And writing nonsense without having the slightest clue about theorist and matstat is not necessary.
And, by the way, Lyapunov's limit theorem has nothing to do with SB.
In general, I would recommend that dads and the rest of the old people who populate the forum attach the state to their posts.
So everyone can see the IQ of this elderly child.
This forum will have some much-needed fun and banter.
It would also be nice to have a note from the psycho-psycho and narcological dispensary in your profile.
And, by the way, Lyapunov's limit theorem has nothing to do with SB.
:))) You're on a roll today, my friend. I think it's time to get out of here - they'll crush your intellect ....