Predicting fumonisins in Iowa corn: Gradient boosting machine learning

IF 2.2 4区农林科学 Q3 CHEMISTRY, APPLIED

Cereal Chemistry Pub Date : 2024-08-12 DOI:10.1002/cche.10824

Emily Branstad-Spates, Lina Castano-Duque, Gretchen Mosher, Charles Hurburgh Jr., Kanniah Rajasekaran, Phillip Owens, H. Edwin Winzeler, Erin Bowers

{"title":"Predicting fumonisins in Iowa corn: Gradient boosting machine learning","authors":"Emily Branstad-Spates, Lina Castano-Duque, Gretchen Mosher, Charles Hurburgh Jr., Kanniah Rajasekaran, Phillip Owens, H. Edwin Winzeler, Erin Bowers","doi":"10.1002/cche.10824","DOIUrl":null,"url":null,"abstract":"<div>\n \n \n <section>\n \n <h3> Background and Objectives</h3>\n \n <p>Fumonisin (FUM), a secondary metabolite from <i>Fusarium</i> spp., poses major concerns for the United States corn industry. This study evaluated a prepublished Illinois-centric predictive model with historical Iowa FUM contamination data using gradient boosting machine (GBM) learning and compared influential predictors with an Iowa-centric model. Corn samples (<i>n</i> = 529) were collected from 2010, 2020, and 2021 in Iowa's 99 counties, and 2011 data were used for independent validation (<i>n</i> = 89).</p>\n </section>\n \n <section>\n \n <h3> Findings</h3>\n \n <p>Applying a 2 ppm (mg/kg) threshold for FUM high and low contamination events, the overall accuracy was 71.08% and 85.39% for the Illinois- and Iowa-centric models in 2011. Balanced accuracies were 60.23% and 50.00% for the Illinois- and Iowa-centric models. For Iowa's remaining years (testing data), the overall accuracy was 98.10%, and balanced accuracy was 50.00%.</p>\n </section>\n \n <section>\n \n <h3> Conclusions</h3>\n \n <p>FUM-GBM analyses determined the top influential predictor for the Illinois-centric model was satellite-acquired normalized difference vegetation index (NDVI) (Veg_index) in March, whereas the top predictor for the Iowa-centric model was precipitation (PRCP) in October.</p>\n </section>\n \n <section>\n \n <h3> Significance and Novelty</h3>\n \n <p>Results indicate that meteorological and agronomic events, such as PRCP and Veg_index in early planting stages and during harvest, may influence the probability of high FUM levels in corn.</p>\n </section>\n </div>","PeriodicalId":9807,"journal":{"name":"Cereal Chemistry","volume":"101 6","pages":"1261-1272"},"PeriodicalIF":2.2000,"publicationDate":"2024-08-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1002/cche.10824","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Cereal Chemistry","FirstCategoryId":"97","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/cche.10824","RegionNum":4,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"CHEMISTRY, APPLIED","Score":null,"Total":0}

引用次数: 0

Abstract

Background and Objectives

Fumonisin (FUM), a secondary metabolite from Fusarium spp., poses major concerns for the United States corn industry. This study evaluated a prepublished Illinois-centric predictive model with historical Iowa FUM contamination data using gradient boosting machine (GBM) learning and compared influential predictors with an Iowa-centric model. Corn samples (n = 529) were collected from 2010, 2020, and 2021 in Iowa's 99 counties, and 2011 data were used for independent validation (n = 89).

Findings

Applying a 2 ppm (mg/kg) threshold for FUM high and low contamination events, the overall accuracy was 71.08% and 85.39% for the Illinois- and Iowa-centric models in 2011. Balanced accuracies were 60.23% and 50.00% for the Illinois- and Iowa-centric models. For Iowa's remaining years (testing data), the overall accuracy was 98.10%, and balanced accuracy was 50.00%.

Conclusions

FUM-GBM analyses determined the top influential predictor for the Illinois-centric model was satellite-acquired normalized difference vegetation index (NDVI) (Veg_index) in March, whereas the top predictor for the Iowa-centric model was precipitation (PRCP) in October.

Significance and Novelty

Results indicate that meteorological and agronomic events, such as PRCP and Veg_index in early planting stages and during harvest, may influence the probability of high FUM levels in corn.

Abstract Image

查看原文本刊更多论文

预测爱荷华州玉米中的伏马菌毒素：梯度提升机器学习

背景与目标烟曲霉毒素（FUM）是镰刀菌属的一种次级代谢产物，是美国玉米产业的主要问题。本研究利用梯度提升机器（GBM）学习，评估了预先公布的以伊利诺伊州为中心的预测模型和爱荷华州的 FUM 污染历史数据，并将有影响力的预测因子与以爱荷华州为中心的模型进行了比较。研究结果以 2 ppm (mg/kg) 为 FUM 高、低污染事件阈值，2011 年以伊利诺伊州和爱荷华州为中心的模型的总体准确率分别为 71.08% 和 85.39%。以伊利诺伊州和爱荷华州为中心的模型的平衡准确率分别为 60.23% 和 50.00%。结论FUM-GBM 分析表明，对以伊利诺伊州为中心的模型而言，影响最大的预测因子是 3 月份卫星获取的归一化差异植被指数（NDVI）（Veg_index），而对以爱荷华州为中心的模型而言，影响最大的预测因子是 10 月份的降水量（PRCP）。重要意义和新颖性结果表明，气象和农艺事件，如播种早期和收获期间的PRCP和Veg_index，可能会影响玉米出现高FUM水平的概率。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Cereal Chemistry 工程技术-食品科技

CiteScore

5.10

自引率

8.30%

发文量

110

审稿时长

3 months

期刊介绍： Cereal Chemistry publishes high-quality papers reporting novel research and significant conceptual advances in genetics, biotechnology, composition, processing, and utilization of cereal grains (barley, maize, millet, oats, rice, rye, sorghum, triticale, and wheat), pulses (beans, lentils, peas, etc.), oilseeds, and specialty crops (amaranth, flax, quinoa, etc.). Papers advancing grain science in relation to health, nutrition, pet and animal food, and safety, along with new methodologies, instrumentation, and analysis relating to these areas are welcome, as are research notes and topical review papers. The journal generally does not accept papers that focus on nongrain ingredients, technology of a commercial or proprietary nature, or that confirm previous research without extending knowledge. Papers that describe product development should include discussion of underlying theoretical principles.