数据子空间提取方法研究及应用

Study and Application of Data Subspace Extraction Method

价值密度低的工业数据面临着如何提取有价值的数据子空间问题,文章应用CART和双CART方法来解决机加工关键尺寸质量参数控制优化问题。该方法还能识别异常数据产生的最大风险路径,并以此提出改善措施用于工艺改善验证。应用实例结果表明,双CART方法用于提取数据子空间可以有效识别异常数据产生的影响因子。

The industrial data with low value density face on the problem of extracting valuable data subspace.The paper studies the CART and Dual-CART methods to solve the control and optimization problem of machining key dimension quality parameters.It can identify the maximum risk path of abnormal data,put forward improvement measures and implement improvement.The application case results show that the Dual-CART method used to extract data subspace can effectively identify the influence factors of abnormal data.