Difference between revisions of "Can we get bacteria to solve a problem large enough to challenge a person?"

From GcatWiki
Jump to: navigation, search
 
(3 intermediate revisions by the same user not shown)
Line 1: Line 1:
If bacteria can be used to optimize, that is minimize or maximize a function, then we are looking at a significant contribution.  Even if the bacteria can get reasonably good solutions to a function, i.e. a heuristic approach, that is significant.  For example, consider the following very small "fitness" function where the variables ''a, b'' and ''c'' are either 0 or 1.
+
If bacteria can be used to optimize, that is minimize or maximize a function, then we are looking at a significant contribution.  Even if the bacteria can get reasonably good solutions to a function, i.e. a heuristic approach, that is significant.  For example, consider the following very small "fitness" function where the variables ''a, b'' and ''c'' are either 0 or 1 (on or off).
  
minimize 2 - a + 2*b*c - b
+
maximize 2 + a - 2*b*c + b
  
This is an example from a small max 2-sat problem [[where all the logical constraints have been translated into a fitness function]].  What values of ''a, b'' and ''c'' minimize this function?  Did you consider this a challenging problem?
+
This is an example from a small max 2-sat problem where all the logical constraints have been translated into a fitness function.  What values of ''a, b'' and ''c'' maximize this function?  Did you consider this a challenging problem? (ans: a = b = 1, c = 0 for an objective value of 4)
  
Is there a way to program e. coli to evolve according to their fitness as measured by this equationNote that a minimization is easily transposed to a maximization via the negation of the function, if that helps the biologists.
+
Is there a way to program e. coli to evolve where their fitness is measured by this or some other function?   
 +
 
 +
Is there a way to try, in parallel, the many possible solutions (different values for the 2^n combinations) and choose the one that has maximized the given function?

Latest revision as of 02:49, 19 May 2009

If bacteria can be used to optimize, that is minimize or maximize a function, then we are looking at a significant contribution. Even if the bacteria can get reasonably good solutions to a function, i.e. a heuristic approach, that is significant. For example, consider the following very small "fitness" function where the variables a, b and c are either 0 or 1 (on or off).

maximize 2 + a - 2*b*c + b

This is an example from a small max 2-sat problem where all the logical constraints have been translated into a fitness function. What values of a, b and c maximize this function? Did you consider this a challenging problem? (ans: a = b = 1, c = 0 for an objective value of 4)

Is there a way to program e. coli to evolve where their fitness is measured by this or some other function?

Is there a way to try, in parallel, the many possible solutions (different values for the 2^n combinations) and choose the one that has maximized the given function?