主成分分析
稳健性(进化)
预处理器
降维
拉曼光谱
计算机科学
人工智能
可解释性
聚类分析
模式识别(心理学)
判别式
材料科学
背景(考古学)
显微镜
分光计
光学
生物系统
化学
物理
地质学
生物
古生物学
生物化学
基因
作者
Sofia Pessanha,Iulian Otel,Valentina Vassilenko,João Silveira,Paulo A. Ribeiro
摘要
Abstract This paper addresses the challenge posed by the size of certain objects that do not conform to the constraints of microscope‐coupled Raman spectrometers, making sample collection impossible due to their inherent value or nature. Specifically, materials like hydroxyapatite‐based substances used in artistic and ornamental carvings, such as bone or ivory, fall within this problematic category. The focus of this study is the enhancement of analytical capabilities in the context of large objects using a Raman microscope system. The study details the innovation involving a remote probe integrated with fiber optics, elaborating on the design and performance aspects, and emphasizing the appropriateness of chosen components in the analysis of ivory artifacts belonging to private collectors. In order to assess the robustness of our discriminative approaches, an archaeological bone and the exposed dentine in a human tooth were also evaluated and compared. Results showed that using an 805‐nm longpass dichroic mirror successfully directed the near‐infrared laser onto the samples and significantly suppressed the Rayleigh scattering contribution to the spectrum. Regarding the preprocessing methods to spectra evaluation essayed, the most promising approach was the use of principal component analysis for dimension reduction followed by k‐means cluster analysis. By leveraging the complementary strengths of PCA and k‐means clustering, the robustness and interpretability of clustering analyses are enhanced.
科研通智能强力驱动
Strongly Powered by AbleSci AI