Evaluation of Pseudorandom Number Generators Based on Residue Arithmetic in Differential Evolution.
ADVANCES IN INTELLIGENT NETWORKING AND COLLABORATIVE SYSTEMS, INCOS-2017(2018)
摘要
Sequences of pseudorandom numbers, produced by various pseudorandom number generators (PRNGs), are essential for the implementation of stochastic algorithms, simulations, and methods including nature-inspired computation. The non-determinism, observed in nature, is emulated with the help of pseudorandom values, predefined probabilities and thresholds, and the properties of generated sequences may have significant impact on the computation. They affect both, the ability of the algorithms to find high-quality problem solutions and the speed of the computation. Residue number systems provide means for alternative representation of integers that is suitable for parallel systems. In this work, several PRNGs based on residue arithmetic and residue number systems are empirically investigated as a source of stochasticity for the differential evolution algorithm.
更多查看译文
关键词
Differential evolution,Pseudorandom numbers,Residue arithmetic,Inversive congruential generator
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络