应用量子粒子群算法优化神经网络的数据库重复记录检测  

Using Quantum Particle Swarm to Optimize Database Duplicate Record Detection Based on Neural Network

作　　者：徐亮 XU Liang(Information Department, Lushan Rehabilitation and Recuperation Center of PLA, Jiujiang 332000, China)

机构地区：[1]解放军庐山康复疗养中心信息科,江西九江332000

出　　处：《微型电脑应用》2022年第1期142-144,149,共4页Microcomputer Applications

摘　　要：神经网络的连接阈值以及权值直接影响数据库重复记录的检测效果,当前方法无法找到最优的神经网络的连接阈值和权值,导致数据库重复记录检测偏差比较大,并且数据库重复记录检测效率低,为了获得更优的数据库重复记录检测结果,提出了量子粒子群算法优化神经网络算法的数据库重复记录检测方法。首先分析当前数据库重复记录检测研究进展,并提取数据库重复记录检测特征向量,然后采用BP神经网络对数据库重复记录检测训练样本进行学习,采用量子粒子群算法确定最优连接阈值和权值,建立理想的数据库重复记录检测模型,最后进行了数据库重复记录检测仿真实验。结果表明,这种方法是一种准确率高、效率高的数据库重复记录检测方法,数据库重复记录检测效果要明显优于其它方法。The threshold and weight of neural network directly affect the detection effect of repeated records in database.The current methods cannot find the optimal threshold and weight of neural network,which leads to a large deviation in the detection of database duplicate records,and the detection efficiency of database duplicate records is low.In order to obtain better detection results of repeated records in database,quantum particles are proposed.Group algorithm optimizes the database duplicate record detection of neural network algorithm.Firstly,the current research progress of database duplicate record detection is analyzed,and the feature vector of database duplicate record detection is extracted.Then,BP neural network is used to learn the training samples of database duplicate record detection,and quantum particle swarm optimization algorithm is used to determine the optimal connection threshold and weight,and an ideal database duplicate record detection model is established.Finally,the database duplicate record detection is carried out by the detection simulation experiment.The results show that this method is a high accuracy and efficiency method of database duplicate records detection,and the detection effect of database duplicate records is obviously better than other methods.

关 键 词：数据库记录 重复检测 连接权值 量子粒子群算法 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]