摘要
根据复杂电磁环境下通信联络的特点,基于大规模通信数据的涌现,采用一种能够在大量干扰和背景信号中筛选出有用信息的算法。通过对覆盖算法的研究,利用概率理论对其算法进行了优化和改进,为应用覆盖算法构建了一个实用的有限混合的概率模型,而后采用最大似然的原理进行优化计算,实现覆盖算法的优化实用化。将这种数据筛选算法应用到受到干扰条件下短波通信的海量信息的挖掘中,取得了良好的效果,验证了其有效性。
Under Complicated Electromagnetic Environment, aiming at the characteristic of the data captured in the commu nication countermeasure, from the view of largescale data mining, the thesis analyzes the traditional neural networks and points out that we can realize data mining by applying covering algorithm to artificial neural networks. The thesis proposes an improved method. Covering algorithm is firstly extended to kernel covering algorithms and we secondly construct a kind of fi nite mixture probabilistic model based on kernel covering algorithms according to the probability meaning of Gaussian func tion and finally introduce the global optimizing computation by "maximum likelihood theory", realize the global optimization problem of the covering algorithm so as to improve its precision and make it suit to largescale data mining. The result of the experiment is given as an example to illustrate the validation of the method.
出处
《指挥控制与仿真》
2013年第3期39-41,55,共4页
Command Control & Simulation
关键词
复杂电磁环境
覆盖算法
通信干扰
概率模型
数据挖掘
complicated electromagnetic environment
covering algorithm
communication interference probabilistic model
data mining