Malaysian Medicinal Leaf Images

Leaf images sample for Malaysian Medicinal Plant classification. Previous study using this dataset are using randomly selected training (45 images) and testing (20 images). Images are pre-processed and numeric values using shape-based features are extracted. Preliminary study using several classifiers such as SMO, J48, NB, k-NN and Random Forest show that SMO produced 50% accuracy. Citation Request Sainin, M. S., & Alfred, R., & Ahmad, F. (2021). Ensemble meta classifier with sampling and feature selection for data with imbalance multiclass problem. Journal of Information and Communication Technology, 20(2), 103-133. https://doi.org/10.32890/jict2021.20.2.1用于马来西亚药用植物分类的叶片图像样本集。过往采用该数据集的相关研究均随机选取45张图像作为训练集、20张图像作为测试集。图像已完成预处理，并提取了基于形状特征的数值化特征。采用SMO、J48、朴素贝叶斯（NB）、k近邻（k-NN）以及随机森林（Random Forest）等多种分类器开展的初步研究显示，SMO的分类准确率仅为50%。 引用请求 Sainin, M. S.、Alfred, R.、Ahmad, F.（2021）. 面向不平衡多分类问题数据的集成元分类器结合采样与特征选择方法. 信息与通信技术期刊, 20(2), 103-133. https://doi.org/10.32890/jict2021.20.2.1