Probability assessment is purely mathematical - page 12

New comment
 
Prival:

5 points
in sync:)))
 

offtop

Please share a working link to matcad or your distr.

 
sever30:

oftop

Please share a working link to matcad or your distr.

rutracker dot org:)))

just don't forget to buy the full version from the manufacturer!!!

 

Unfortunately, I've got Win7 -64 and can't get the matcad on it. version 15 has already been released, but it won't work for me ((

http://rutracker.org/forum/viewtopic.php?t=3030331

 

Acquire? :) and without acquiring?

 
alsu:
About presence/absence of dependencies I agree. But about differentiation I would argue: each differentiation operation nullifies one order of dependence if represented polynomially. So even if we get that there is no dependence in the differentiated series, it doesn't mean that there wasn't one in the original series.


I'm not suggesting to differentiate N times, I'm suggesting just once (i.e. we need to analyse the increments). On the whole - I agree with you.

The ACF of the increments will be _like_ the delta function. However, correlation coefficients lying in the interval [-2/sqrt(n); 2/sqrt(n)] (which are usually considered insignificant) may well be significant for incremental series with long-run memory.

 
sever30:

Acquire? :) and without acquiring?

and without acquiring - write your own matcad:)))) or download from a tracker and turn yourself in to the militia:)))
 
alsu:
Without buying, write your own matcad:)))) or download from a tracker and turn yourself in to the police:)))

thank you kind man:)
 
lea:


I am not suggesting to differentiate N times, but only once (i.e. I need to analyse the increments). On the whole - I agree with you.

The ACF of the increments will be _like_ the delta function. However, correlation coefficients lying in the interval [-2/sqrt(n); 2/sqrt(n)] (which are usually recognized as insignificant) may well be significant for increments of series with long-run memory.

I couldn't agree more. But it is worth discussing the question of how useful it is in practice .
 
lea:


...However, correlation coefficients lying in the interval [-2/sqrt(n); 2/sqrt(n)] (which are usually considered insignificant) may well be significant for series increments with long-term memory.

Regarding increments, this is of course my opinion, but a lot of people here talk about price increments, substituting this notion with Close bar increments. Which from my point of view is not quite correct. Most probably, we should analyze the increment of this point (asc+bid)/2, this point is closer to the notion of price, at leastthe floating spread will have less influence.

This can only be done by analyzing ticks, bars won't do the trick. But that's just my opinion...

Hint, where did this formula come from,

interval [-2/sqrt(n); 2/sqrt(n)]

I'm just curious, I think I calculated it somehow differently, if need be I can dig around and find it.

1...5678910111213141516171819
New comment