变精度粗糙集挖掘在丙酮精制中的应用

The Application of Data Mining based on the Variable Precision Rough Set Theory in Acetone Refining

在标准粗糙集数据挖掘算法基础上,研究变精度粗糙集数据挖掘方法在丙酮精制产品质量预报中的应用,并与标准粗糙集方法、BP算法及Apriori算法进行比较。采用实际丙酮精制生产数据进行实验,结果表明,变精度粗糙集数据挖掘算法的应用效果显著优于BP算法和Apriori算法,相对于标准粗糙集方法也有明显提高,具有一定的实用价值和研究前景。

Based on rough set data mining algorithm, data mining model based on the variable precision rough set theory is applied to prediction of product qualities in Acetone refining process. The method is also compared with back-propagation neural network （BPNN）, Apriori algorithm and original rough set algorithm. The experimental results on the real industrial data of a Acetone refining process demonstrate that the variable precision rough set data mining method achieves better performance than BPNN and Apriori algorithm, and also better than rough set algorithm.