Evolutionary Computing in R with the ecr package

4:55 pm - 5:20 pm 24 Thursday

Track

Evolutionary Computation Software

Evolutionary algorithms are efficient general purpose stochastic optimization procedures, which evolve solutions for specific problems based on the Darwinian principle of evolution theory. Different evolutionary algorithms mainly distinguish themselves by the applied building blocks, i.e., variation and selection operators as well as problem representation and fitness function.We introduce the R package ecr, short for evolutionary computing in R. The package offers a comprehensive collection of building blocks for the construction of evolutionary algorithms for both single- and multi-objective optimization problems in the famous statistical programming language R. An extensible plugin system allows for the easy creation of evolutionary problems solvers build up of predefined or even self-implemented evolutionary operators. Besides a set of popular evolutionary multi-objective algorithms like NSGA-II and SMS-EMOA is included. This work introduces the package, describes the main underlying principles and gives some introductory examples on how to use it. 

Speaker:
Jakob Bossek