Number series recognition software?

  • Context: Undergrad 
  • Thread starter Thread starter fenian
  • Start date Start date
  • Tags Tags
    Series Software
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
4 replies · 12K views
fenian
Messages
3
Reaction score
0
Does anybody know if there is any software available that is capable of recognizing patterns in a series of numbers (and possibly predicting the next number to appear)?

Here's an example of what I'm looking for.

0111100100110000001011001101011010101000110010110101010110100010011001110100110101000100010101100000

I, personally, do not recognize any pattern in that sequence, whatsoever, but there there is one. This much I do know.

I have been looking all over the internet, trying to find something that will determine that kind of pattern, but have had no luck thus far. Maybe I'm just typing in the wrong thing to search for? I don't know, but if anybody has any suggestions for me, I would greatly appreciate it.
 
Mathematics news on Phys.org
That problem strikes me as computationally intractable. What would the program do? Check through a list of every conceivable relationship between elements of the string? Even if a string did have a pattern, it could very easily be visible only over the long term (i.e. millions of characters).

How do you know that there is a pattern, exactly? Humans beings are horrible judges; we tend to see patterns everywhere, even in completely random data (blame evolution). That sequence looks pretty much like what you would expect a random sequence to look like.
 
fenian said:
Does anybody know if there is any software available that is capable of recognizing patterns in a series of numbers (and possibly predicting the next number to appear)?

Here's an example of what I'm looking for.

0111100100110000001011001101011010101000110010110101010110100010011001110100110101000100010101100000

I, personally, do not recognize any pattern in that sequence, whatsoever, but there there is one. This much I do know.

I have been looking all over the internet, trying to find something that will determine that kind of pattern, but have had no luck thus far. Maybe I'm just typing in the wrong thing to search for? I don't know, but if anybody has any suggestions for me, I would greatly appreciate it.

You could easily write a little script to look up any sequence here ...

http://oeis.org/
 
Already looked it up broken into 1, 2, and 3 sets a bit ago... although each one is 50 binary digits long, which makes it look like it might not be binary, and definitely has to be paired in 1,5,10,25... so...

Broken up into length 5 segments, translated from binary doesn't exceed 26. Although the 0 at the end would give you a 27th character... :D

15,4,24,2,25,21,21,8,25,13 10,26,4,25,26,13,8,17,11,0
 
Last edited:
Thanks a lot for your help guys. I think Matt Benesi may be on to something with the speak of binary.

It may have been helpful earlier to state that this sequence could probably be looked at very similar to digital where 0=off and 1=on. If you look at it that way, I guess I'd say I'm trying to figure out where the 1's would appear if continuing that sequence. I'll see if I can manage to figure it out with the observation that each 5 length segment does not exceed 26.