摘要
总结了稳健主成分分析、稳健主成分回归、稳健偏最小二乘回归和稳健连续回归等各种稳健算法的新近成果.研究表明,稳健算法可以检测并规避异常值的影响.稳健算法应用红外光谱分析中可望优化定性、定量预测模型.
The recent achievements of sundry robust algorithms, including robust principal component analysis, robust principal component regression, robust partial least square regression, robust continuum regression, etc. are summarized. Robust algorithms can detect outliers and circumvent their negative impacts. The application of robust algorithms in infrared spectral analysis can be expected to improve the qualitative and quantitative models.
出处
《分析测试技术与仪器》
CAS
2013年第2期71-76,共6页
Analysis and Testing Technology and Instruments
基金
中国科学院知识创新重要方向项目(KZCX2-YW-QN411)资助
关键词
稳健算法
异常值
红外光谱分析
robust algorithm
outlier
infrared spectral analysis