Sample screening and modeling method based on spectrum similarity and chemical index difference

• Main Authors: ZHONG YONGJIAN, GUO MENGHAO, TONG SHEN, TIAN YUNONG, LI HAIFENG, WU LINGTONG, ZHENG HONGBIN, HAO XIANWEI, LIAO FU, GE CHUAN, BI YIMING, WANG ZHULIN, XIA JUN, LIU JIANGUO, ZHANG ZHAOPENG, WU JIZHONG, WANG HUI, LI SHITOU
• Format: Patent
• Language: Chinese; English
• Published: 28.01.2022
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; HANDLING RECORD CARRIERS; INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIRCHEMICAL OR PHYSICAL PROPERTIES; MEASURING; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS; TESTING

The invention discloses a sample screening and modeling method based on spectrum similarity and chemical index difference. The relation between the most similar sample pair in a sample set and the chemical index difference is investigated. The difference range of chemical values of two samples under high similarity is obtained by measuring the similarity of the samples. As the similarity value is reduced, the maximized difference degree of the chemical indexes is increased, the samples with the trend are effective samples, and the relation between the spectrum and the chemical indexes can be well established no matter what modeling mode is adopted. If the trend cannot be embodied, the batch of data lacks the distinction degree between the spectrum and the chemical indexes, so that the modeling cannot be successful. Therefore, the method is helpful for users to timely eliminate factors such as improper modeling methods and fewer modeling samples, and clearly give whether the batch of data can be modeled or not