Variable selection in a GPU cluster using delta test

Authors:
A. Guillén;M. van Heeswijk;D. Sovilj;M. G. Arenas;L. J. Herrera;H. Pomares;I. Rojas
Affiliations:
Department of Computer Architecture and Computer Technology Universidad de, Granada, Spain;Department of Information and Computer Science, Aalto University School of Science, Finland;Department of Information and Computer Science, Aalto University School of Science, Finland;Department of Computer Architecture and Computer Technology Universidad de, Granada, Spain;Department of Computer Architecture and Computer Technology Universidad de, Granada, Spain;Department of Computer Architecture and Computer Technology Universidad de, Granada, Spain;Department of Computer Architecture and Computer Technology Universidad de, Granada, Spain
Venue:
IWANN'11 Proceedings of the 11th international conference on Artificial neural networks conference on Advances in computational intelligence - Volume Part I
Year:
2011

Citing 11
Cited 0

Efficient and Accurate Parallel Genetic Algorithms

Efficient and Accurate Parallel Genetic Algorithms
Heterogeneous computing and parallel genetic algorithms

Journal of Parallel and Distributed Computing - Problems in parallel and distributed computing: Solutions based on evolutionary paradigms
Output value-based initialization for radial basis function neural networks

Neural Processing Letters
Finding the embedding dimension and variable dependencies in time series

Neural Computation
Boosting the Performance of a Multiobjective Algorithm to Design RBFNNs Through Parallelization

ICANNGA '07 Proceedings of the 8th international conference on Adaptive and Natural Computing Algorithms, Part I
Parallel multiobjective memetic RBFNNs design and feature selection for function approximation problems

Neurocomputing
Minimising the delta test for variable selection in regression problems

International Journal of High Performance Systems Architecture
Effective input variable selection for function approximation

ICANN'06 Proceedings of the 16th international conference on Artificial Neural Networks - Volume Part I
Improving the performance of multi-objective genetic algorithm for function approximation through parallel islands specialisation

AI'06 Proceedings of the 19th Australian joint conference on Artificial Intelligence: advances in Artificial Intelligence
Gradual distributed real-coded genetic algorithms

IEEE Transactions on Evolutionary Computation
Parallelism and evolutionary algorithms

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

The work presented in this paper consists in an adaptation of a Genetic Algorithm (GA) to perform variable selection in an heterogeneous cluster where the nodes are themselves clusters of GPUs. Due to this heterogeneity, several mechanisms to perform a load balance will be discussed as well as the optimization of the fitness function to take advantage of the GPUs available. The algorithm will be compared with previous parallel implementations analysing the advantages and disadvantages of the approach, showing that for large data sets, the proposed approach is the only one that can provide a solution.