Multi expression programming

Multi Expression Programming (MEP) is a genetic programming variant employing a linear representation of solutions. MEP representation is inspired by Three-address code. MEP introduces a unique feature: the ability to encode multiple solutions, of a problem, in the same chromosome. In this way one can explore larger zones of the search space. For most of the problems this advantage comes with no running-time penalty compared with genetic programming variants encoding a single solution in a chromosome.

Example of MEP program

Here is a simple MEP program:

1: a
2: b
3: + 1, 2
4: c
5: d
6: + 4, 5
7: * 3, 5

On each line we can have a terminal or a function. In the case of functions we also need pointers to its arguments.

When we decode the chromosome we obtain multiple expressions:

E1 = a,

E2 = b,

E4 = c,

E5 = d,

E3 = a + b.

E6 = c + d.

E7 = (a + b) * d.

Which expression will represent the chromosome? In MEP each expression is evaluated and the best of them will represent the chromosome. For most of the problems, this evaluation has the same complexity as in the case of encoding a single solution in each chromosome.

See also

• Genetic programming

External links

• Multi Expression Programming website