IN-SEASON PREDICTION OF POTATO PRODUCTION USING UAV-BASED HYPERSPECTRAL IMAGERY AND MULTI-TASK MACHINE LEARNING FOR SUSTAINABLE IRRIGATION MANAGEMENT

The overall objective of this proposal is to develop a multi-task machine learning model for in-season prediction of potato production under different irrigation rates. Specific objectives for the proposed research are as follows:1) Conduct field research trials and UAV surveys for model development.2) Develop feature extraction algorithms for time series hyperspectral data fusion.3) Develop a multi-task machine learning model for in-season prediction of yield and quality traits, by investigating the underlying cross-relatedness between different traits.