Algorithms For Dummies

Solving Problems Quickly

As machine learning takes off and we count more and more on computers to solve 

problems, the issue of how quickly a computer can solve a problem becomes criti-

cal. The P versus NP problem simply asks whether a computer can solve a problem 

quickly when it can verify the solution to the problem quickly. In other words, if 

the computer can reasonably ascertain that a human response to a problem is cor-

rect in polynomial time or less, can it also solve the problem itself in polynomial 

time or less?

This  question  was  originally  discussed  in  the  1950s  by  John  Nash  in  letters  to  

the National Security Agency (NSA) and again in letters between Kurt Gödel and 

John  von  Neumann.  In  addition  to  machine  learning  (and  AI  in  general),  this 

 particular  problem  is  a  concern  to  many  other  fields,  including  mathematics, 

cryptography, algorithm research, game theory, multimedia processing, philosophy, 

and economics.

---

404

 

   

  PART 6