Genetic programming: on the programming of computers by means of natural selection
Genetic programming: on the programming of computers by means of natural selection
An introduction to genetic algorithms
An introduction to genetic algorithms
Proceedings of the 11th Annual conference on Genetic and evolutionary computation
Incorporating expert knowledge in evolutionary search: a study of seeding methods
Proceedings of the 11th Annual conference on Genetic and evolutionary computation
Solving iterated functions using genetic programming
Proceedings of the 11th Annual Conference Companion on Genetic and Evolutionary Computation Conference: Late Breaking Papers
Programming Massively Parallel Processors: A Hands-on Approach
Programming Massively Parallel Processors: A Hands-on Approach
On GPU based fitness evaluation with decoupled training partition cardinality
EvoApplications'13 Proceedings of the 16th European conference on Applications of Evolutionary Computation
| Hi-index | 0.00 |
Gene expression programming (GEP) is a functional genotype/phenotype system. The separation scheme increases the efficiency and reliability of GEP. However, the computational cost increases considerably with the expansion of the scale of problems. In this paper, we introduce a GPU-accelerated hybrid variant of GEP named pGEP (parallel GEP). In order to find the optimal constant coefficients locally on the fixed function structure, the Method of Least Square (MLS) has been embedded into the GEP evolutionary process. We tested pGEP using a broad problem set with a varying number of instances. In the performance experiment, the GPU-based GEP, when compared with the traditional GEP version, increased speeds by approximately 250 times. We compared pGEP with other well-known constant creation methods in terms of accuracy, demonstrating MLS performs at several orders of magnitude higher in terms of both the best residuals and average residuals.