Shuichi Shinmura Boeken

Focusing on the first discriminant theory of linearly separable data, the book presents Theory3, which builds on previous theories and utilizes 169 microarrays for analysis. It emphasizes the importance of accurate diagnoses by addressing misclassified patients within medical data. The author introduces RIP, an optimal-linear discriminant function designed to minimize misclassifications, showcasing its effectiveness in distinguishing between cases. This work aims to enhance the understanding and application of discriminant analysis in medical contexts.

The book uniquely compares eight linear discriminant functions (LDFs) across various datasets, including Fisher's iris data and medical data with collinearities. It introduces a 100-fold cross-validation method tailored for small samples and presents a straightforward model selection procedure to identify the optimal model based on minimum M2. The Revised IP-OLDF, evaluated using the MNM criterion, demonstrates superior performance compared to other M2s across the examined datasets, making it a significant contribution to statistical modeling and data analysis.