6 计算智能1.pptx

计算智能 (Computational Intelligence，CI)Artificial Neural NetworksFuzzy logicEvolutionary computation中南大学 刘丽珏贝兹德克(J.C. Bezdek)What is CI?The first published definition According to James C. Bezdek(1994)一个计算智能系统应当只涉及数值(低层)数据，含有模式识别部分，不应用人工智能意义上的知识，而且能够呈现出:计算适应性 (Computational adaptivity )计算容错性 (Computational fault tolerance )接近人的速度 (Speed approaching human-like turnaround ）近似于人的误差率（Error rates that approximate human performance ）…” 当一个智能计算系统以非数值方式加上知识，即成为人工智能系统Although used fairly widespread, there is no commonly accepted definitionWhat is CI?计算智能是信息科学与生命科学相互交叉的前沿领域，是现代科学技术发展的一个重要体现Umbrella term used to coverFuzzy Logic(模糊逻辑)Neural Nets(神经网络)Evolutionary Algorithms(进化计算)要定义“计算智能”并不简单。要在一个正式的定义中容纳诸如模糊集合、神经网络、进化计算、机器学习、贝叶斯推理等各自具有其既定特性的不同领域，即便不是不可能，也是非常困难的。What is CI?Some other opinions:Conference: Computational Intelligence - Methods Applications- CIMA2005Defining Computational Intelligence is not straightforward. It is difficult, if not impossible, to accommodate in a formal definition disparate areas with their own established individualities such as fuzzy sets, neural networks, evolutionary computation, machine learning, Bayesian reasoning, etc.Book: “Computational Intelligence: An Introduction”, Andries P. Engelbrecht, Wiley 2002Computational intelligence is the study of adaptive mechanisms to enable or facilitate intelligent behavior in complex and changing environments. As such, computational intelligence combines artificial neural networks, evolutionary computing, swarm intelligence and fuzzy systems.计算智能是关于在复杂和变化的环境中如何实现智能行为的自适应机制研究。因此，计算智能综合了人工神经网络、进化计算、群智能和模糊系统。艾伯哈特(Eberhart)贝兹德克(J.C. Bezdek)AI vs. CI绝大多数的AI/CI 研究者将这两者看成不同的研究领域CI 构成了AI的一个不同选择 (R.C.Eberhart)目标相近CI深入系统的核心AI 位于系统的外层AI 包含了CI (Bezdek)AI vs. CIDue to BezdekThree levels of system complexityA: Artificial or SymbolicB: Biological or OrganicC: Computational or Numeric system计算智能是一种智力方式的低层认知，它与人工智能的区别只是认知层次从中层下降至低层而已。中层系统含有知识，低层系统则没有InputComplexityOutputHuman Knowledge(＋)Sensory Data BNNBPRBIB－OrganicKnowl