Formation of International Research Unit for Prediction Agriculture based on Spatiotemporal Multidimensional Data

• Overview

  Mechanistic physical models are not suitable for analysis based on incomplete information and are labor-intensive and time-consuming to obtain the necessary data for parameterization. Data-driven models such as machine learning can be a useful tool in such a case. The objective of this research unit is to establish an international prediction agriculture research platform for the prediction of various environmental events related to agriculture using mathematical models. The unit will further promote research on quantitative prediction of resource dynamics, ecosystem structure and dynamics, and biological effects in agricultural land and surrounding environment using mathematical models, and will focus on (1) the improvement of existing models, (2) the integration of mechanistic physical models and data-driven models, and (3) the collection of training data for building predictive data-driven models.

  

Members

Hirotaka Saito (Institute of Agriculture / Professor)
Junko Nishiwaki (Institute of Agriculture / Associate Professor)