All Results

Subjects: Computer Science >> Other Disciplines of Computer Science

目前诸多模式识别任务的识别精度获得不断提升，在一些任务上甚至超越了人的水平。单从识别精度的角度来看，模式识别似乎已经是一个被解决了的问题。然而，高精度的模式识别系统在实际应用中依旧会出现不稳定和不可靠的现象。因此，开放环境下的鲁棒性成为制约模式识别技术发展的新瓶颈。实际上，在大部分模式识别模型和算法背后蕴含着三个基础假设：封闭世界假设、独立同分布假设、以及大数据假设。这三个假设直接或间接影响了模式识别系统的鲁棒性，并且是造成机器智能和人类智能之间差异的主要原因。本文简要论述如何通过打破三个基础假设来提升模式识别系统的鲁棒性。 |

Entropy-partition of Complex Systems and Emergence of Human Brain’s Consciousness

Xi GuangchengSubjects: Mathematics >> Statistics and Probability

The three frameworks for theories of consciousness taken moust seriously by neuroscientists are that consciousness is a biological state of the brain,the global workspace perspective,and the perspective of higher state. Consciousness is discussed from viewpoint of theory of Entropy—partition of complex system in present article. Human brain’s system self-organizably and adaptively implements partition、aggregation and integration, and consciousness emerges. |

submitted time
2019-01-10
Hits*15795*，
Downloads*1637*，
Comment
*0*

Subjects: Computer Science >> Integration Theory of Computer Science

针对特定领域内自动化识别既有概念和发现新概念的问题，提出了一种基于条件随机场和信息熵的抽取方法。通过使用条件随机场对文本中的概念词进行边界预测，与词典中的概念对比，筛选出新概念的候选项并找出其大概位置，然后由互信息和左右熵分别判断概念窗口内的概念内部结合度和概念边界自由度，从而发现新的专业概念。实验表明，使用该方法进行概念发现比单独使用条件随机场的方法有更好的效果，基于字和词的模型概念发现的准确率分别提升了20.06%和46.54%。 |

submitted time
2019-01-03
From cooperative journals:《计算机应用研究》
Hits*659*，
Downloads*381*，
Comment
*0*

Gibbsian representation of knowledge in infinite dimensional random neural networks(IDRNN)

xi guangchengSubjects: Mathematics >> Statistics and Probability

Abstract. In studying of a class of random neural network, some of relative researchers have proposed Markov model of neural network. Wherein Markov property of the neural network is based on “assuming”. To reveal mechanism of generating of Markov property in neural network, it is studied how infinite-dimensional random neural network (IDRNN) forms inner Markov representation of environment information in this paper.Because of equivalence between markov property and Gibbsian our conclusion is that knowledge is eventually expressed by extreme Gibbs probability measure—ergodic Gibbs probability measure in IDRNN. This conclusion is also applicable to quantum mechanical level of IDRNN. Hence one can see “ concept “- “ consciousness” is generated at particle(ion) level in the brain and is experienced at the level of the neurons; We have discussed also ergodicity of IDRNN with random neural potential. |

submitted time
2018-11-07
Hits*15665*，
Downloads*1233*，
Comment
*0*

Using associative neural network to interpret syndromes of Traditional Chinese Medicine

Xi Guangcheng Chen Jianxin Yi JianQiangSubjects: Mathematics >> Modeling and Simulation

Millions of people benefit form Traditional Chinese Medicine TCM every day. Unfortunately till now TCM has not been accepted as science by world especially western people. Bian Zheng Lun Zhi is distillation of TCM. Syndrome is key in system of Bian Zheng Lun Zhi. Study about the syndrome is core of study of basic theory of TCM. We creatively interpret TCM through a view of cognitive science and take syndromes of TCM as concepts of brain. This paper try to introduce syndrome to western people in order to let western people understand our viewpoints more easily the best method is to adopt a manner that is easily understood by them already exists and has been thought to be right. So we employ neural network presented by foreign people as brain model instead of network presented by us Using two classic case of TCM we successfully clarify the three main properties of syndrome in TCM. |

submitted time
2017-05-25
Hits*4300*，
Downloads*1998*，
Comment
*0*

[1 Pages/ 5 Totals]