Peking University, April 15, 2017: During April 2 to 4, 2017, PKU College of Engineering Professor Shi Leyuan’s group won the worldwide second award in the Syngenta Crop Challenge held by 2017 INFORMS Conference on Business Analytics & Operations Research.
The sponsor of this competition is Syngenta, an international leading agricultural technology company. Since September 1, 2016, more than 600 teams throughout the world,includingHarvard University, Stanford University etc.,have submitted research plans, suggesting the competitionisvery fierce. Through a sever two-round selection by review committee, five teams entered the final on February 24, 2017, and competed ultimately in Las Vegas, United States on April 3, 2017.
In the competition, the title of paper submitted by Professor Shi Leyuan’s team is A Decision Making Approach for Soy Seed Variety Selection via Hedging Against Weather Risk. They put forward an innovative idea of comprehensively using big data analysis technology and deep neural networks to construct agricultural output prediction model. The optimal solution strategy based on mixed integer quadratic programming was obtained from the model, creatively solving the problem of optimal ration for crop seed selection.The original points of this scheme is using covariance matrix to hedge risks caused by weather uncertainty. Simultaneously develops decision analysis support tools aiming at the accurate prediction of national grain output, and seed preparation optimal strategy by the combination of big data technology, deep learning technology and optimization theory and methods, which is fairly propagable. The INFORMS review committee gave consistent credit for the solution scheme. Robin Lougee, chairman of review committee and industry research chief scientist of IBM Research and T.J. Watson Research Center, and Joseph Byrum, chief scientist of Syngenta company, pointed out thatit had great theoretical value and practical significance.
In recent years, Professor Shi Leyuan’s teamhas tightly combined their researchwith the national development strategic demand, undertaking many important projects, including The National Natural Science Foundation major project “Collaborative management theory of high end equipment manufacturing network based on Internet and big data environment”and The National Nonprofit Industry Special Research “Analysis of the data storage platform for big data environment”. They also carried out fruitful researches in fields such as “Internet+” intelligent manufacturing, agricultural big data analysis and optimization, which are highly admitted by domestic and foreign counterparts and industrial circles.
The study was collaborated by researchers from Peking University and University of Wisconsin-Madison (UWM) and sponsored by The National Natural Science Foundation and The National Nonprofit Industry Special Research.
INFORMS Conference on Business Analytics & Operation Research is held by INFORMS annuallyand is a prestigiousconference in the field of global operations research and business analysis. The conference aims at applying advanced operations, management science, big data, and machine learning technologies to solve practical problems in business analysis. Thousands of experts, scholars from the domain of global operations, management science and business analysis, and enterprise elites attended the conference.
Written by: Zhang Jingwen
Edited by: Wang Chengsiou
