Efficient Learning of Site-Specific Management in Precision Agriculture
Presented by:
Yuji Saikai
Department of Agricultural & Applied Economics
University of Wisconsin - Madison
Wednesday, February 27, 2019
12:00 pm-1:30 pm
Taylor-Hibbard Seminar Room (Rm103)
In precision agriculture (PA), farmers face new opportunities and challenges. They can achieve greater efficiency by optimizing site-specific management based on unique information on each site obtained through various sensor technologies. At the same time, since modern PA equipment often requires substantial investment, without sufficient return generated by effective use of the technology, PA may merely increase financial risks and even lower profitability. We propose a model specifically designed for profit-maximization in PA as well as a general yet practical solution algorithm. The model is based on a class of optimization techniques called Bayesian optimization, which is known for its sample efficiency and capability to handle complex objective functions. A simulation result indicates the effectiveness of our approach to unleashing the potential of PA.