Fitness first and fatherless crossover

Fitness first and fatherless crossover
复制标题

健身第一与无父交叉

DOI:
10.1145/3449726.3459437
复制
发表时间:
2021
期刊:
--
影响因子:
--
通讯作者:
Langdon W
Langdon W
中科院分区:
--
文献类型:
--
作者:
Langdon W

文献摘要

参考文献

被引文献

相似文献

当项目非常大,并且适应度评估非常快时(例如,由于只评估项目的一小部分),在创建项目之前评估适应度是有效的,因为只需要创建有孩子的个体。节省可以达到50%,即使在高度融合的人群中,也可以节省13.5%。在GP交叉中,一个孩子是由两个父母创造的,但是捐赠的父母(妈妈)贡献的远远超过另一个(爸爸)。来自父类的子树通常很小,可以从每个父类中提取并在交叉之前保存。这就产生了单亲杂交,当它与适合度结合在一起时,进一步减少了杂交的数量。即使在最坏的情况下,减少也是exp(-1) = 37%。对于大型树,即使在相似适应度的种群中,消除单身汉和老处女也是可行的,并且可以减少运行时间和内存消耗。与通常的M+2N相比,(N)个人口M的多线程实现中的存储约为0.63M + N,在实践中节省≥17%。在Intel i7-9800X 16线程的3.8GHz台式机(CPU带宽85 GByte/秒)上,我们实现了每秒6920亿GP操作,692gb gpop。
When programs are very large and fitness evaluation is fast (e.g. due to evaluating only a tiny fraction of the program) it can be efficient to evaluate fitness before creating the program, as only individuals with children need be created. The saving can be > 50% and even in highly converged populations, it savese-2= 13.5%.In GP crossover one child is created from two parents but the root donating parent (mum) contributes far more than the other (dad). The subtree from the father is usually small and can be extracted from each dad and saved before crossover. This gives single parent crossover, which when combined with fitness first, further reduces the number of crossovers. Even in the worst case the reduction is exp(-1) = 37 percent.With large trees, even in populations of similar fitness, eliminating bachelors and spinsters is feasible and can reduce both runtime and memory consumption. Storage in a (N) multi-threaded implementation for a population M is about 0.63M + N, compared to the usual M+2N, in practice saving ≥ 17%. We achieve 692 billion GP operations per second, 692 giga GPops, on an Intel i7-9800X 16 thread 3.8GHz desktop (CPU bandwidth 85 GByte/second).
用于世代遗传编程的 C 语言多线程内存高效交叉
DOI: --
发表时间: 2020
期刊: ACM SIGEVOlution
影响因子: --
作者:
W. Langdon
通讯作者: W. Langdon
DOI: --
发表时间: 2019
期刊: Artificial Life Conference Proceedings
影响因子: --
作者:
W. Langdon;W. Banzhaf
通讯作者: W. Banzhaf
DOI: --
发表时间: 2021
期刊: European Conference on Genetic Programming
影响因子: --
作者:
W. Langdon
通讯作者: W. Langdon