Scholarlink

<p>Common bean is a crop of significant socioeconomic importance in many developing countries and plays a key role in food security. Therefore, identifying excellent genotypes for grain yield is a crucial step in breeding programs. This study aimed to evaluate the performance of artificial neural networks in predicting common bean genotypes considering different grain yield ranges, using data obtained from multiple environments. The model was trained and tested with phenotypic variables to classify genotypes into grain yield categories (poor, medium, good, and excellent). Results demonstrated satisfactory overall performance, with an overall accuracy of 70%, and greater discriminative capacity for the extreme classes, especially excellent and poor. The area under the curve reinforced the model’s effectiveness, with values above 0.90 for the excellent and poor classes, which are central to genotype recommendation and discard decisions. Overall, neural networks proved to be a promising tool for supporting decision-making in common bean breeding, enabling efficient identification of promising genotypes and elimination of less productive ones during the final field evaluation stage.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Neural Networks Applied to Plant Breeding for Predicting Grain Yield in Common Bean Genotypes

  • 期刊论文

  • 发表日期：

    2026年2月27日

查看全文
  • Luan Tiago dos Santos Carbonari,
  • Carlos Joaquin Zacarias Junior,
  • Marissa Prá de Souza,
  • Charlene Barboza Bussolaro,
  • Carina Lopes Djadjo,
  • Altamir Frederico Guidolin,
  • Sydney Antonio Frehner Kavalco,
  • Jefferson Luís Meirelles Coimbra

摘要

Common bean is a crop of significant socioeconomic importance in many developing countries and plays a key role in food security. Therefore, identifying excellent genotypes for grain yield is a crucial step in breeding programs. This study aimed to evaluate the performance of artificial neural networks in predicting common bean genotypes considering different grain yield ranges, using data obtained from multiple environments. The model was trained and tested with phenotypic variables to classify genotypes into grain yield categories (poor, medium, good, and excellent). Results demonstrated satisfactory overall performance, with an overall accuracy of 70%, and greater discriminative capacity for the extreme classes, especially excellent and poor. The area under the curve reinforced the model’s effectiveness, with values above 0.90 for the excellent and poor classes, which are central to genotype recommendation and discard decisions. Overall, neural networks proved to be a promising tool for supporting decision-making in common bean breeding, enabling efficient identification of promising genotypes and elimination of less productive ones during the final field evaluation stage.