Plant Breeding Evaluation with Rank Entropy-based Decision Tree

详细信息    查看全文

• 作者：Xiangyu Zhao^* ; ^** ; ^{zhaoxy@nercita.org.cn" class="auth_mail" title="E-mail the corresponding author} ; Zhongqiang Liu^* ; ^*** ; ^{liuzq@nercita.org.cn" class="auth_mail" title="E-mail the corresponding author} ; Fang Dan^* ; ^**** ; ^{danf@nercita.org.cn" class="auth_mail" title="E-mail the corresponding author} ; Kaiyi Wang^* ; ^** ; ^{wangky@nercita.org.cn" class="auth_mail" title="E-mail the corresponding author}
• 关键词：Breeding Evaluation ; Decision Support Systems ; Ordinal Classification ; Rank Entropy ; Crop Phenomics
• 刊名：IFAC-PapersOnLine
• 出版年：2016
• 出版时间：2016
• 年：2016
• 卷：49
• 期：16
• 页码：336-340
• 全文大小：423 K

文摘

With the rapid development of breeding equipments, large scale breeding becomes possible, massive data evaluation has to be done by software automatically. In this circumstance, this paper tries to evaluate variety with the collected informative data through data mining technologies. According to the characteristics, plant breeding evaluation is considered as an ordinal classification task, and a rank entropy-based decision tree algorithm is proposed to do this classification. The algorithm uses historical trait phenotype and evaluation data to construct decision trees, which are then used to generate evaluations for future cultivars with their trait phenotype. To demonstrate the effectiveness, experiments are carried out on three groups of soybean variety comparison tests (early-maturity, medium-maturity, and green soybean). This work can free breeders from a large number of basic work while increase the efficiency of plant breeding.