BACKGROUND Horse gram (Macrotyloma uniflorum (Lam.) Verdc.) is an underutilized pulse crop with high drought resistance traits and rich source of protein. But, conventional breeding method for high yielding and abiotic stress tolerant germplasm is hampered by the scarcity of morphological data sets. Therefore, classification of Horse gram adapted to various agro-ecological zones prevailing various stress factors to exhibit homogenous genotype. Nowadays, several machine learning (ML) methods were used in the field of plant phenotyping. RESULTS Here we adopted unsupervised learning techniques of K-means clustering algorithm for their usefulness to analyze six important morphological traits such as plant shoot length, total plant height, flowering percentage, number of pods per plant, pod length, number of seeds per plant and seed length variants between germplasm. Unsupervised clustering revealed that, twenty germplasm accessions were grouped in four clusters in which high yielding trait was predominantly observed in the cluster 2. CONCLUSION Therefore, these findings could guide ML based classification easily to characterize the suitable germplasm on the basis of high yielding variety for the different agro-ecological zones. This article is protected by copyright. All rights reserved.

中文翻译：

---

耐旱马革兰种质的形态特征：通过机器学习进行分类

背景马革（Macrotyloma uniflorum (Lam.) Verdc.）是一种未被充分利用的豆类作物，具有高度抗旱性状和丰富的蛋白质来源。但是，传统的高产和非生物胁迫耐受种质育种方法受到形态学数据集稀缺性的阻碍。因此，马氏菌的分类适应了各种农业生态区，存在各种胁迫因素，表现出均一的基因型。如今，在植物表型分析领域中使用了几种机器学习（ML）方法。结果 在这里，我们采用 K-means 聚类算法的无监督学习技术来分析六种重要的形态性状，如植物芽长、植物总高度、开花百分比、每株豆荚数、豆荚长度、每株植物的种子数和种质之间的种子长度变异。无监督聚类显示，20 个种质材料被分为四个聚类，其中在聚类 2 中主要观察到高产性状。 结论 因此，这些发现可以指导基于 ML 的分类，以根据高产品种对合适的种质进行表征不同的农业生态区。本文受版权保护。版权所有。这些发现可以指导基于 ML 的分类，以根据不同农业生态区的高产品种轻松表征合适的种质。本文受版权保护。版权所有。这些发现可以指导基于 ML 的分类，以根据不同农业生态区的高产品种轻松表征合适的种质。本文受版权保护。版权所有。