Potato Biodiversity: Classifying 50 Genotypes Using Linear Discriminant Analysis

This research paper investigates potato biodiversity by employing linear discriminant analysis (LDA) to classify fifty genotypes based on their nutritional and physicochemical characteristics. LDA is a multivariate statistical method that identifies the discriminatory power of a set of variables (in this case, the nutritional and physicochemical composition of genotypes) for grouping (in this case, the genotypes). By calculating linear discriminant functions, samples can be assigned to the optimal classification, minimizing within-group variation and maximizing between-group differences. In this study, researchers collected samples from fifty different potato genotypes and measured their nutritional and physicochemical composition, such as protein content, starch content, mineral content, etc. They then applied LDA to determine the ability of these variables to assign samples to their correct genotype. Through calculating linear discriminant functions, they could then assign new samples to the optimal classification. The objective of this classification process is to identify differences between genotypes and find the nutritional and physicochemical characteristics associated with these differences. By doing so, researchers can understand the potential variations among different genotypes and provide information regarding nutritional and physicochemical characteristics for selecting specific genotypes or optimizing potato varieties.