Genetic Programming
Genetic Programming
复制标题
DOI：
10.1007/bfb0055923
复制
发表时间：
1998
期刊：
影响因子：
--
通讯作者：
Moshe Sipper
中科院分区：
文献类型：
--
作者：
Moshe Sipper
进化计算（EC）为当今计算机科学带来了巨大的希望。在20世纪50年代早期开始之后，它被少数科学家追求，直到20世纪80年代它作为一个快速增长的领域起飞。演化计算方法解决了各种应用领域中的问题，这些问题可以被转换为抽象的搜索空间，由这些问题定义，由搜索过程遍历。EC所有分支的共同特点是它们都是面向路径的搜索方法，即访问并存储各种候选解作为进一步探索的起点。勘探主要通过随机方法进行，尽管确定性移动也是可能的。该领域的先驱之一汉斯·布雷默曼（Hans Bremermann）在1962年写道：“致力于解决问题、定理证明和模式识别的各个小组的经验似乎都指向同一个方向：这些问题很坚韧。似乎没有一条皇家或一个简单的方法可以一下子解决我们所有的问题。涉及大量可能性的问题无法通过纯粹的数据处理量来解决。我们必须追求质量，追求精致，追求技巧，追求我们能想到的每一种独创性。比今天的计算机更快的计算机将有很大的帮助。我们会需要他们的。然而，当我们从原则上考虑问题时，今天的计算机大约和将来一样快。“大自然提供了丰富的技巧和改进的来源。正是在这种精神下，EC诞生了，起源于生物学中的达尔文进化论。种群、突变、重组和选择等生物学概念已被转移到更抽象的计算环境中并投入使用。遗传程序设计（GP）是EC最年轻的分支，自1992年John Koza出版了一本关于该主题的书以来，它得到了迅速的发展。在过去几年中发表了800多篇论文。GP可以被认为是自动编程的少数方法之一，因为正在进化的种群结构是计算机程序。GP已经成功地应用于大量的困难问题，如自动设计，模式识别，机器人控制，神经网络的合成，符号回归，音乐和图片生成，和许多other.This卷包含的程序EuroGP'98，第一次欧洲遗传编程研讨会在巴黎，法国，1998年4月14日至15日举行。EuroGP'98是在欧洲举行的第一次完全致力于遗传规划的活动。其目的是为GP领域的欧洲和非欧洲研究人员以及行业人士提供一个展示其最新研究并讨论当前发展和应用的机会。该讲习班由EvoNet（进化计算卓越网络）赞助，作为EvoGP（进化网络遗传编程工作组）的活动之一。
Evolutionary Computation (EC) holds great promise for computer science today. After an early start in the 1950s, it was pursued by a handful of scientists until it took off as a rapidly growing field in the 1980s. Evolutionary computational approaches solve problems in various application domains which can be cast as abstract search spaces, defined by those problems, to be traversed by search processes. The common feature of all branches of EC is that they are path-oriented search methods, ie, a variety of candidate solutions is visited and stored as starting points for further exploration. Exploration takes place mostly by stochastic means, although deterministic moves are also possible. One of the forefathers of the field, Hans Bremermann, wrote in 1962:" The experiences of various groups who work on problem solving, theorem proving, and pattern recognition all seem to point in the same direction: These problems are tough. There does not seem to be a royal road or a simple method which at one stroke will solve all our problems.... Problems involving vast numbers of possibilities will not be solved by sheer data processing quantity. We must look for quality, for refinements, for tricks, for every ingenuity that we can think of. Computers faster than those of today will be of great help. We will need them. However, when we are concerned with problems in principle, present day computers are about as fast as they will ever be." Nature provides a rich source of tricks and refinements. It is in this spirit that EC came into being, deriving from Darwinian evolution in biology. Biological notions such as population, mutation, recombination, and selection have been transferred and put to use in more abstract computational contexts. Genetic Programming (GP), the youngest branch of EC, has grown rapidly since the publication of a book on the subject by John Koza in 1992. More than 800 papers have been published over the last few years. GP can be considered one of the few methods of automatic programming since the structures of the population being evolved are computer programs. GP has already been applied successfully to a large number of difficult problems like automatic design, pattern recognition, robotic control, synthesis of neural networks, symbolic regression, music and picture generation, and many others.This volume contains the proceedings of EuroGP'98, the First European Workshop on Genetic Programming held in Paris, France, on 14-15 April, 1998. EuroGP'98 was the first event entirely devoted to genetic programming to be held in Europe. The aims were to give European and non-European researchers in the area of GP, as well as people from industry, an opportunity to present their latest research and discuss current developments and applications. The workshop was sponsored by EvoNet, the Network of Excellence in Evolutionary Computing, as one of the activities of EvoGP, the EvoNet working group on genetic programming.