Scholarlink

<p>Efficient in vitro propagation of <i>Phalaenopsis</i> orchids is essential for improving plant quality and production efficiency. This study applied artificial neural networks (ANNs) alongside experimental validation to identify optimal concentrations of two biostimulants, i.e., chitosan and 24-epibrassinolide (24-EBL), for enhancing morphological and photosynthetic performance in <i>Phalaenopsis</i> plantlets. A dataset generated from controlled experiments was utilized to train an ANN with the Levenberg–Marquardt back-propagation algorithm. Model accuracy was confirmed through five-fold cross-validation. Predictions indicated that 12 mg L<sup>-</sup>¹ chitosan combined with 0.3 mg L<sup>-</sup>¹ 24-EBL produced the most favorable growth responses. The optimized treatment yielded remarkable improvements, including a &gt; 700% increase in leaf area and significant gains in number of leaves, root formation, and photosynthetic efficiency. Validation trials supported the ANN predictions, demonstrating the reliability and practical value of this approach. These findings highlight the potential of machine learning to accelerate micropropagation and provide a scalable framework for refining culture conditions in ornamental horticulture.</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

Artificial neural network-based optimization of chitosan and 24-epibrassinolide for enhancing Phalaenopsis plantlet growth in vitro

  • 期刊论文

  • 发表日期：

    2026年5月13日

查看全文
  • Zahra Mahdavi,
  • Shirin Dianati Daylami,
  • Keyvan Asefpour Vakilian,
  • Kourosh Vahdati

摘要

Efficient in vitro propagation of Phalaenopsis orchids is essential for improving plant quality and production efficiency. This study applied artificial neural networks (ANNs) alongside experimental validation to identify optimal concentrations of two biostimulants, i.e., chitosan and 24-epibrassinolide (24-EBL), for enhancing morphological and photosynthetic performance in Phalaenopsis plantlets. A dataset generated from controlled experiments was utilized to train an ANN with the Levenberg–Marquardt back-propagation algorithm. Model accuracy was confirmed through five-fold cross-validation. Predictions indicated that 12 mg L-¹ chitosan combined with 0.3 mg L-¹ 24-EBL produced the most favorable growth responses. The optimized treatment yielded remarkable improvements, including a > 700% increase in leaf area and significant gains in number of leaves, root formation, and photosynthetic efficiency. Validation trials supported the ANN predictions, demonstrating the reliability and practical value of this approach. These findings highlight the potential of machine learning to accelerate micropropagation and provide a scalable framework for refining culture conditions in ornamental horticulture.