Fast Feature Selection in a GPU Cluster Using the Delta Test

Abstract

Feature or variable selection still remains an unsolved problem, due to the infeasible evaluation of all the solution space. Several algorithms based on heuristics have been proposed so far with successful results. However, these algorithms were not designed for considering very large datasets, making their execution impossible, due to the memory and time limitations. This paper presents an implementation of a genetic algorithm that has been parallelized using the classical island approach, but also considering graphic processing units to speed up the computation of the fitness function. Special attention has been paid to the population evaluation, as well as to the migration operator in the parallel genetic algorithm (GA), which is not usually considered too significant; although, as the experiments will show, it is crucial in order to obtain robust results.

The delta test: The 1-NN estimator as a feature selection criterion

“…0140 LARS 1-08 2,3,17,21,35,36,37,39 0. 0131 LARS 1-09 2,3,17,19,21,35,36,37,39 0.0131 LARS 1-10 2,3,17,19,21,33,35,36,37,39 0.0124 LARS 1-11 2,3,17,19,20,21,33,35,36,37,39 0.0100 LARS 1-12 2,3,17,19,20,21,33,35,36,37,39,48 0.0098 LARS 1-13 2,3,17,19,20,21,33,35,36,37,39,44,48 0. 0098 LARS 1-14 2,3,17,19,20,21,33,35,36,37,39,44,48,49 0.…”

“…0097 LARS 1-15 2,3,17,19,20,21,33,35,36,37,39,44,48,49,51 0. 0096 LARS 1-16 2,3,17,19,20,21,33,35,36,37,39 36,38 0.0858 Relief 1-03 36,37,38 0.0505 Relief 1-04 19,36,37,38 0.0327 Relief 1-05 3,19,36,37,38 0.0329 Relief 1-06 3,19,36,37,38,39 0.0263 Relief 1-07 3,19,35,36,37,38,39 0.0196 Relief 1-08 3,19,21,35,36,37,38,39 0.0179 Relief 1-09 3,17,19,21,35,36,37,38,39 0.0167 Relief 1-10 3,17,19,21,…”

“…Experimental evidence comparing to alternative methods is also provided. Choosing an efficient search scheme for high-dimensional tasks is mostly left as a practical matter of implementation, and several papers specifically about optimising the Delta test have also been published in the literature [13], [14], [15], [16], [17], [18], [19].…”

Confluences among Big Data, Finite Element Analysis and High Performance Computing

“…It is currently an opensource system that is often used in conjunction with Hadoop to provide a distributed, scalable and high performance file system in place of HDFS. It enables HPC engines to maintain performance by aggregating multiple I/O paths to multiple servers in the compute cluster (Guillén et al, 2014;Slack, 2014).…”

A Thermodynamical Selection-Based Discrete Differential Evolution for the 0-1 Knapsack Problem

“…Evolutionary algorithms (EAs) are effective heuristic algorithms for solving global optimization problems [7][8][9][10][11]. Moreover, many kinds of evolutionary algorithms have been successfully applied to solve the 0-1 knapsack problem in the past few decades.…”