Application of Biometrics and Biostatistics to Agricultural Science

Statistics and biometrics have become one of the most important disciplines in various fields, not only in agricultural sciences. This is mainly due to the following three factors: (1) Improvements in data measurement techniques have led to the collection of diverse and large amounts of biological and agronomical data that were not collected in the past, (2) Data science methods have been developed to connect and model the collected biological and agronomical data, (3) Computational power to use such methods has improved. These circumstances have made statistical and biometric methods essential tools for extracting knowledge from diverse and voluminous biological and agronomical data.
In this lecture, various types of biological and agronomic data will be used as examples to illustrate the methods used to analyze them. The lecture will be given in a hands-on style using R, Python and Matlab, and is intended to provide students with practical analysis techniques. The first one-third of the entire course is devoted primarily to analytical methods related to summarizing, visualizing, and modeling relationships in multivariate data. In the second one-third of the course, students will learn about linear models, linear mixed models, local regression, and nonlinear models. In the last one-third of the course, students will learn about image analysis, machine learning, and deep learning. Although a wide range of methods from introductory to advanced will be used, emphasis will be placed on acquiring the ability to perform actual analysis on one's own, rather than on explaining the principles of the methods.