【摘要】 研究对新生物数据库中特征目标数据进行有效检测,可有效保障新生物数据库系统的稳定运行。但由于新生物数据库目标数据分布散乱不存在连锁反应,导致传统方法只能检测出一种特征目标数据,存在检测误差大的弊端。提出一种改进多标签分类的新生物数据库中特征目标数据优化检测方法。上述方法先采用K均值聚类对新生物数据库中的数据进行聚类生成均值参考点,对生成的均值参考点进行拟合,确定特征目标数据的区域范围,利用遗传算法搜索特征目标数据的特征子集,将多标签扩展成可以对特征目标数据的特征子集进行评估的多标签,组建特征目标数据优化检测模型,弥补了传统算法只能检测出一种特征目标数据弊端。仿真结果表明,优化方法可以显著地提升新生物数据库中特征目标数据优化检测性能。更多还原

【Abstract】 Research on the effective detection of the feature target data in new biological database can effectively guarantee the stable operation of the new biological database system. In the paper,we proposed a new classification based on improved tabbed biological feature target data in a database optimization method. Whith this method,the author firstly used k- means clustering to generate mean reference for new biological data in the database point and processed fitting for mean reference point to determine the local scope of the characteristic target data by using the genetic algorithm search features of target data subset,to expand more labels to evaluate the feature of target data feature subset and more labels. The optimization model of feature target data was set up,which makes up for the shortcomings of traditional algorithm which can only detect one kind of characteristic target data. Simulation experiments show that an optimization method can significantly improve the new biological optimization detection performance characteristics of the target data in the database.更多还原