Randomization…

After a hectic few days, and some unnecessary forks, I’m back on track again. Anyway’s, I’ve been thinking on some Randomization Techniques ( Links 1, 2, 3…) and recently read some text about them ( Not exactly, “Monte Carlo” by Fisher). There are a number of methods by which one can generate Random Numbers from a Digital Machine, Like 1) Entropy in the Physical Memory, 2) Heat Variations in the Various parts of CPU, HDD, 3) Voltage variations in clock pulses, etc. But most of them are trackable. and after you know which Algorithm is using which techniques, the Random Numbers generated by that algorithm will not be so Random!

Patterns everywhere. Mathematics show that, there’s a certain pattern in everything we can feel, i.e : Matter, energy. But that’s an issue of level of intractability. Like, it depends on the processing power of human minds. Something like NP-hard problems. It’s solvable (“It’s just a problem… ” –Richard Feynman!) yet, it’s far beyond present human intelligence (or may be lack of information).

So the point is, what we can make, is to think up something which is not trackable upto some extent. We may take this level as a vector and maximize it with some human parameters. Sounds AI? yeah. that can be done by some specified mathemetics/geometry, which is not as straightforward as Euclidian Maths/Geometry model. A good help of Fuzzy Logic is also needed, becuse, a level of nondeterministic acceptance wil be there.

This text is a tangled mess! I just got some foolish thoughts and coughed up in here. Maybe someday, if things go right, we’ll come up with a new approach to Randomization and, maybe, Clever decision making capabilities by Machine Intelligence!

Advertisements