1
$\begingroup$

Let me change my question:

I have a sequence of 512 numbers (these 512 numbers can have a value between 0 and 255) (in computer mode these numbers can be between 00000000 and 11111111 (bits)). For these 512 numbers I calculate the average value, standard deviation value and kurtosis value. Now I want to calculate the values for the distribution of averages, distribution of standard deviations and distribution of Byte values. Finaly, I want to compare the distributions from other sequesces of 512 numbers to find out the similarity (as percentage) that two sequences have. I dont understand how can I calculate these distributions.

for example Distribution of Averages is : the probability that an average chosen from all the averages of a memory block is of value B in the range of 0-255.

Dxb=Pr((B+1))>Xj>=B ,Xj=Average.

$\endgroup$
3
  • $\begingroup$ To check I understand your question: you have two sequences of 512 bytes, call them $X$ and $Y$, and you want a measure of how similar they are? $\endgroup$
    – Chris Taylor
    May 19, 2011 at 10:41
  • $\begingroup$ yes exactly! i want to have a percent of their similarity. for example that sequences X and Y are 78% similar. $\endgroup$
    – nomercy
    May 19, 2011 at 10:46
  • $\begingroup$ To say that 'sequences X and Y are 78%' similar, in this approach, makes no sense to me. Before asking "how to calculate" something, you must be clear what results you expect. Say X and Y have the same values in distinct order- are they 100% similar? Say $x_i = y_i+1$ : the have same variance and higher centered moments, and similar average (difference =1) How similar would you call them? $\endgroup$
    – leonbloy
    Jun 24, 2011 at 14:06

1 Answer 1

Reset to default
1
$\begingroup$

For the new question, it appears that you have no problem calculating the mean, standard deviation, and kurtosis of one series. Then if you have a bunch of series and want to calculated the distribution of means you do it the same way. You have data that comes as $(i,mean_i,sd_i,kurt_i)$ and ignore where it came from. So you can calculate the mean and standard deviation of all the means. As you have not defined "the similarity as percentage" I don't know what to do with that.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .