There are multiple source codes in this repository:
mep_symbolic_regression - basic MEP algorithm. Solves symbolic regression and binary classification problems. This should be the starting point for everyone.
mep_binary_classification - basic MEP algorithm. Solves symbolic regression and binary classification problems. This should be the starting point for everyone.
mep_multiple_pop - MEP with multiple (sub)populations for regression and binary classification problems.
mep_multi_class - MEP algorithm. Solves regression and multi-class classification problems.
mep_threads - MEP with multiple (sub)populations evolved in different threads for regression and binary classification problems.
mep_circuits - MEP for evolving digital circuits.
mep_parity_adfs - MEP algorithm with Automatically Defined Functions for Even-Parity problems.
mep_evolve_EAs - MEP for evolving Evolutionary Algorithms.
Take a .cpp program from a folder and compile it. You also need a dataset from the the same folder as the cpp file. Make sure that you specify the path correctly.
There are multiple datasets in this repository:
building1.txt - a symbolic regression problem taken from PROBEN1.
cancer1.txt - a binary classification problem taken from PROBEN1.
iris.txt - a classification problem with 3 classes.
gene1.dt - a classification problem with 3 classes. Format of output is 1-of-m.
2x2_multiplier.txt, 3x3_multiplier.txt - this is a problem of designing digital circuits (for 2x2 and 3x3 multiplication).
even_6_parity.txt - even parity problems with 7 inputs.