论文部分内容阅读
在现实中,许多数据库都是动态变化的,非增量约简方法处理这些数据需要花费大量的时间和空间。增量技术是处理动态数据的有效方法。首先介绍了计算知识粒度的增量机制,然后提出了基于知识粒度的增量约简算法,当一些对象增加到决策表时,能够利用原有决策表的知识粒度和约简,快速计算出增加对象后的知识粒度和约简,并通过理论分析验证了增量方法可以减少计算属性约简的时间复杂度,最后用增量方法和非增量方法对UCI数据集进行一系列试验。试验结果表明,所提增量算法在处理动态数据时能够节省大量的计算时间。
In reality, many databases are dynamic, non-incremental reduction method to deal with these data takes a lot of time and space. Incremental technology is an effective way to handle dynamic data. Firstly, the incremental mechanism of calculating the granularity of knowledge is introduced. Then, an incremental reduction algorithm based on knowledge granularity is proposed. When some objects are added to the decision table, the knowledge granularity and reduction of the original decision table can be used to quickly calculate the incremental objects After the knowledge granularity and reduction, and through theoretical analysis verify that the incremental method can reduce the time complexity of computing attribute reduction, and finally a series of experiments UCI dataset using incremental methods and non-incremental methods. The experimental results show that the proposed incremental algorithm can save a lot of computing time when dealing with dynamic data.