Molecular imprinted polymer (MIP)-based electrodes have been utilized for monitoring the variations in catechin and epigallocatechin gallate (EGCG) concentrations at different stages of black tea processing. These are two particular flavanol variants present in tea. Differential pulse voltammetry technique was used to assess the electrochemical behavior of both the sensors in each step of tea processing. A standard arrangement involving three electrodes was used to perform the experiments. The sensor response at various tea processing stages was effectively captured. The acquired data were first preprocessed and then different clustering techniques, such as K-means clustering, FCM clustering, and agglomerative clustering, were used to segregate the data at each stage based on the catechin and EGCG contents. These methods were able to develop proper clusters with varying levels of precision. The silhouette scores for different steps are 0.643, 0.642, 0.744, 0.509, and 0.482, respectively. A silhouette score close to 1 signifies that clusters have been configured more perfectly.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Discrimination of Catechin and EGCG Through Various Tea Processing Stages Using Clustering Techniques

  • Debanjana Ghosh,
  • Shreya Nag,
  • Debangana Das,
  • Runu Banerjee Roy

摘要

Molecular imprinted polymer (MIP)-based electrodes have been utilized for monitoring the variations in catechin and epigallocatechin gallate (EGCG) concentrations at different stages of black tea processing. These are two particular flavanol variants present in tea. Differential pulse voltammetry technique was used to assess the electrochemical behavior of both the sensors in each step of tea processing. A standard arrangement involving three electrodes was used to perform the experiments. The sensor response at various tea processing stages was effectively captured. The acquired data were first preprocessed and then different clustering techniques, such as K-means clustering, FCM clustering, and agglomerative clustering, were used to segregate the data at each stage based on the catechin and EGCG contents. These methods were able to develop proper clusters with varying levels of precision. The silhouette scores for different steps are 0.643, 0.642, 0.744, 0.509, and 0.482, respectively. A silhouette score close to 1 signifies that clusters have been configured more perfectly.