Discussion of article "Exploring Seasonal Patterns of Financial Time Series with Boxplot" - page 22
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
Are there module names for Python or R? Python is better.
I don't express myself in pythons :-)
just the sum/concurrence of the modulus...how can I put it more precisely.... Zk[i]=X[i+k]+X[i+k*2]... i.e. in a row, sum the multiples of k. If there is indeed a cycle close to k, the minimum and maximum will be clearly marked in the convolution.
they are like this:
the closer you get to the sine, the more accurate the period and the tempo... the whole method is like a "focus pointing".
I don't express myself in pythons :-)
just sum/multiplication by modulus... how can I put it more precisely.... Zk[i]=X[i+k]+X[i+k*2]... i.e. in a series, you sum up multiples of k. If there is indeed a cycle close to k, then the minimum and maximum will be clearly marked in the convolution.
they are like this:
the closer you get to the sine, the more accurate you get the period and the tempo. the whole method is a bit like "focusing".
Oops. Can I get two more of these? If they're in a row, no offence.
I don't express myself in pythons :-)
just sum/multiplication by modulus... how can I put it more precisely.... Zk[i]=X[i+k]+X[i+k*2]... i.e. in a series, you sum up multiples of k. If there is indeed a cycle close to k, then the minimum and maximum will be clearly marked in the convolution.
they are like this:
the closer you get to the sine, the more accurate the period and the tempo... the whole method is like a "focus pointing".
I still have a vague understanding
In general, just look at the dependencies (ratios) of points in adjacent boxplots, if there is something inside there, you can pull out more signals, but more fitting as a consequence
Well, something like that, yeah. The idea in my head was to just find a regression of one set of points on another and see the relationships. That's the easiest thing I could think of. Then you could get additional signals. I don't know how statistically correct that would be.
I.e. dependencies caused by changes in the previous hour will be traded, not only in the current hour. From the behaviour in the previous hour, the signals in the current hour will vary, if a correlation can be found. This is closer to fitting or optimisation.It is quite a normal idea. You can draw and calculate) Draw - scatter diagram, calculate - correlation coefficient and its significance.
It's quite a normal idea. You can draw and calculate) Draw - scatter diagram, calculate - correlation coefficient and its significance.
Actually, let's call this process "double charge bomb" and apply it to pair trading as well
If such a theme will suit everyone, I will do it
And the bomb with triple will be on machine learning.Very interesting article. Maxim, thank you.
You're welcome!
Actually, let's call this process the "double charge bomb" and apply it to pair trading as well
If this topic is acceptable to everyone, I'll do it
I'm all for it.
Oops. Can I have two more of these plots? If it's in a row, no offence.
on the picture inside for more than a year..this is the convolution on which the grid is drawn Gana (that's specifically now scares everyone away)
I'll give it out tomorrow, today under the evening and beer already no, really lazy :-) to run a script, collect data, then recalculation and drawing. But I can chat :-)
It's a chicken and egg question. You can convince yourself of the correctness of any approach.
From my point of view, you have done implicit optimisation. Any study is implicit optimisation, which is always a subset of explicit optimisation.
Non-optimisation is the absence of a statistical study. Roughly speaking, when you made a hypothesis without data and it was confirmed.
optimisation
statistical study
implicit optimisation is a subset of explicit optimisation.