二维云模型及其在预测中的应用

PLANAR MODEL AND ITS APPLICATION IN PREDICTION

本文在一维正态云的基础上，进一步提出了二维正态云的思想，给出了描述二维云的数字特征和二维云的数学模型，探讨了二维云发生器的构成方法和应用前景．尤其是在基于数据库的知识发现方面，在定性和定量相结合的基础上，为进行多种预测，融随机性与模糊性为一体，表达不确定性，提供了更系统、更高层的工具．

Based on three technical mainstays, artificial intelligence, statistics and probability, database, a new research area, known as data mining and knowledge discovery (DMKD), has been established and has made a rapid development which offers lots of practical application systems aiding partial automation of data mining in large real databases. A new mathematical representation of linguistic concepts,which is called clouds model (CM), together with its theory, is presented in the paper. With the new model of clouds and a virtual rule engine, a novel uncertainty reasoning mechanism is proposed, and is used in the process of mining predictive knowledge. Based on the data clouds model, the concept of planar clouds is raised,as a more complex tool, it integrates fuzziness and randomness in an inseparable way. The digital characteristics of a planar clouds is represented by expected value Ex, entropy En and hyper entropy H. In the paper, the method of describing planar clouds is given, and the algorithm of the clouds is analyzed deeply, at last the application of planar clouds generators is discussed.