摘要
类似正态分布在实际的生活与生产中分布最为广泛,精确确定的模糊概念隶属函数严重影响该类数据的预测精度。云模型把随机性和模糊性结合起来,用数字特征期望、熵和超熵,揭示随机性与模糊性的关联性。基于正态云模型设计预测算法,放宽形成正态分布要求的前提条件,把精确确定隶属函数放宽到构造正态隶属度分布的期望函数,更简单、直接地完成类正态分布的数据的预测,因而更具有普遍适用性。
Normal distribution is the most common probability distribution function, and is widely used in natural science and social science, but in fact there are so many distributions which are not the precise normal distribution, but are only similar to normal distribution, accu- rately established subordination fuzzy function seriously reduces the prediction accuracy of such data. Cloud model combines randomness and fuzziness well, it reveals the association of randomness and fuzziness with digital feature expectation, entropy and ultra-entropy. The prediction algorithm designed on normal cloud model basis relaxes the prerequisite of forming a normal distribution and widens the accurately determined subordination function to expectation function which forms normal subordination distribution, thus it is more universal, easily and directly completes the prediction of normal distribution. It provides the more accurate prediction data for the national life and the production.
出处
《计算机应用与软件》
CSCD
2009年第9期78-79,105,共3页
Computer Applications and Software
基金
山东省自然科学基金项目(Y2007G65)
关键词
类正态分布
云模型
数字特征
隶属度
预测
Similar normal distribution Cloud model Digital feature Subordination Prediction