A Common Lisp framework for experimenting with meta-optimizing semantic evolutionary search (MOSES) and related approaches to learning with probability distributions over program spaces based on:

• reduction to normal form
• representation-building
• deme (sub-population) management
• probabilistic model-building (aka estimation-of-distribution algorithms)

Contributing

• Join the plop-dev mailing list
• File a feature request or bug under Issues, or tackle one already listed (also see the ToDo list)
• Try the code on your favorite Common Lisp implementation other than CLISP & SBCL and see if it works or not
• Add comments to the code, critique the design, add unit tests ...
• Try the system out for supervised classification / timeseries prediction / whatever on your favorite data set (note that it will probably do rather badly at the moment as the system is still very immature and some important subcomponents are just stubs) and post your results