Implementation of the Vector Evaluated Genetic Algorithm (VEGA). VEGA should
be avoided in practice, since many modern algorithms outperform it and
exhibit better convergence properties, but is included due to its historical
significance. VEGA is considered the earliest MOEA. It supports M
objectives during the selection phase by selecting M different subgroups,
each selected based on the i-th objective value, for i=1,...,M.
There is one small algorithmic difference between this implementation and
[1]. In [1], applying the genetic operators fills the entire population.
However, since custom variation operators can be specified, it is possible
that the population will not be filled completely. As a result, this
implementation will continue selecting parents until the population is full.
References:
Schaffer, D. (1985). Multiple Objective Optimization with Vector
Evaluated Genetic Algorithms. Proceedings of the 1st International
Conference on Genetic Algorithms, pp. 93-100.