Syngenta and the Analytics Society of INFORMS are proud to recognize a team from Iowa State University as the winner of the 2020 Syngenta Crop Challenge in Analytics.
The Syngenta Crop Challenge in Analytics is an annual collaborative effort between Syngenta and the Analytics Society of the Institute for Operations Research and the Management Sciences (INFORMS). Analytics and data science play a vital role in agriculture, when farmers are facing increasing pressures from climate change, soil erosion and biodiversity loss, and from consumers’ changing tastes in food.
This year’s competition invited experts in data analytics, mathematics and statistics to use real-world agriculture data to construct a model that can predict the performance of crossing any two inbred hybrid lines. Understanding how corn plants react when facing certain stresses can be a powerful tool for developing hybrids, allowing breeders to focus on the best possible combinations to meet grower needs.
The winning team, which included Javad Ansarifar, Faezeh Akhavizadegan and Lizhi Wang, was awarded a $5,000 prize for the submission, “Yield Performance of Plant Breeding Prediction with Interaction Based Algorithm.”
They represented Iowa State University, which is in Ames, Iowa. The group collaborates extensively with operations researchers, statisticians, agronomists, breeders, farmers and agriculture industry experts.
“Over the past several years, the Syngenta Crop Challenge has been a fascinating venue to accelerate innovation in plant science,” says Javad Ansarifar of the Iowa State University team. “Making the right crosses is crucial in plant breeding to continuously improve crop performance. Our model helps breeders make the most promising crosses without having to rely on large-scale trial-and-error. This work is part of our research effort in designing explainable artificial intelligence in agriculture.”
Hosted by INFORMS, the leading international association for operations research and analytics professionals, the competition concluded during a virtual awards ceremony held on April 30. Five finalist teams presented their submissions for evaluation by the prize committee.
“We received a large number of high-quality submissions for the 2020 Syngenta Crop Challenge in Analytics, and all the finalists stood out with creative approaches and thorough data analysis to address the challenge,” says Durai Sundaramoorthi, area coordinator and senior lecturer of data analytics at Olin Business School, Washington University in St. Louis, and Crop Challenge prize committee chair. “After careful deliberation, we selected the team from Iowa State University as the winner due to their balanced approach to solve the problem with both accuracy and an interpretable model.”
The runner-up submission, “Hybrid Crop Yield Prediction Using Deep Factorization Methods with Integrated Modeling of Implicit and Explicit High-Order Latent Variable Interactions,” authored by Shouyi Wang, Jie Han, Fangyun Bai and Ho Manh Lin from University of Texas at Arlington received a $2,500 prize.
The third-place entry, “Combining Strong Learners to Predict Yield of Maize Hybrids,” authored by Craig A. Rolling, Isaac Akogwu, Christopher Cotter and Yalda Zare from Benson Hill in St. Louis received a $1,000 prize.
“The need for continued innovation in agriculture is urgent, and data analytics play an important role in helping us meet the needs of a growing population,” says Gregory Doonan, head of advanced analytics, Syngenta, and Crop Challenge judge. “We are committed to bringing innovation to farmers faster to help improve grower profitability and increase the sustainability of agriculture. All the finalists demonstrated unique approaches to address the complexities of crop breeding and advance crop productivity.”
