#18 Add additional problems and Precode implementations

There are numerous other problems that can be explored
using genetic programming besides the current example:
Artificial Ant. As new ideas are mentioned, please add them
to the list. If you would like to work on an idea, add it as a
separate tracker item containing details of your
implementation ideas.

A few examples from Chapter 7 of Koza’s book "Genetic
Programming" (1992):
- Cart Centering (applying a unit force to center a cart over a
point in one dimension)
- Symbolic Regression (discovering a mathematical formula
to fit a given set of two-dimensional or higher data)

Both of these examples require a new Precode
implementation, namely a Floating-Point Precode that uses
hierarchies of functions accepting and returning floating-point
values. I.E. it would allow items like "sin( cos( x ), 5 + y )".
The single-precision "float" data type would probably be
sufficient for working with these problems, provided the range
of data values in the Environment is reasonable.