移动终端智能产品服务需求数据集成获取仿真

Mobile Terminal Intelligent Product Service Demand Data Integration Acquisition Simulation

对智能产品服务需求数据的集成获取是实现移动终端个性化应用推送的关键技术.智能产品服务需求数据具有复杂性和多样性,当前方法无法对其进行准确、有效集成获取,存在着内存占用大、运行时间长、集成获取精度低等问题,针对以上问题,提出一种基于遗传算法的移动终端智能产品服务需求数据集成获取方法.采用小波变换对移动终端智能产品服务需求数据进行预处理,将产品服务需求数据时间序列转换到频谱空间中,得到此空间内距离最小、类间集成最大的变换基系数;基于遗传算法建立数据集成获取模型,将提取的交换基函数作为需求数据特征输入,计算产品服务需求数据聚焦在多维属性空间内的聚类中心,将训练集与所属的类别进行关联,获得产品服务需求数据的属性集分类和信息增益,并根据信息增益值完成数据集成获取.实验结果表明,所提方法的数据集成获取精度可达98.53%,内存占用及运行时间均优于当前方法,具有较好的应用价值.

The current method has some problems such as large memory occupation, long running time and low integration accuracy. Due to above problems, this paper puts forward a method for acquiring the service requested data of mobile terminal intelligent product based on genetic algorithm. At first, the wavelet transform was used to preprocess the intelligent product service requested data of mobile terminal, and then the time series of product service requested data was converted into the spectrum space, so as to obtain the transform radix with the smallest distance and the largest inter-class integration. Moreover, the model of acquiring data integration was built based on the genetic algorithm, and then extracted exchange basis function was inputted as the demand data feature. In addition, the product service demand data was calculated to focus on the cluster center in multi-dimensional attribute space. Finally, the training set was associated with the category to obtain the classification of attribute set of product service requested data and the information gain. Thus, data integration acquisition based on the information gain value was completed. Simulation results prove that the data integration accuracy of proposed method can reach 98.53%. Meanwhile, the memory occupation and running time are better than that of current method, which has good application value.