| 基于有监督与半监督模型的茶叶产地溯源研究 |
| |
| 引用本文： | 戴柯磊,杨科莹,周子懿,段赟杰,洪雪珍.基于有监督与半监督模型的茶叶产地溯源研究[J].现代农业科技,2021,24(24). |
| |
| 作者姓名： | 戴柯磊 杨科莹 周子懿 段赟杰 洪雪珍 |
| |
| 作者单位： | 中国计量大学,中国计量大学,中国计量大学,中国计量大学,中国计量大学 |
| |
| 基金项目： | 浙江省公益研究计划项目(LGN19C160004)、杭州市农业与社会发展科研项目(20191203B26)、国家创新创业训练项目(202010356015)。 |
| |
| 摘 要： | 茶叶市场掺假现象愈演愈烈，为了保护消费者权益，促进茶叶产业的发展，对茶叶进行快速高效的溯源鉴定是尤其重要的。目前，市面上采用的茶叶溯源方法大多需要数量较为固定的样本量去进行预测分析，但在实际应用中，由于大规模样本的需求，会导致预测成本进一步提高。本文以浙江安吉白茶为研究对象，通过在不同的有监督与半监督模型下对不同样本量预测准确度的比较，为实际应用中，在保证预测准确度的前提下，合理经济化的选取样本量提供了参考意见。结果显示，在有标签样本较少的情况下，基于半监督算法建立的茶叶溯源模型效果比有监督算法建立的茶叶溯源模型效果更好。
|
| 关 键 词： | 安吉白茶 近红外光谱 产地溯源 样本量选取 |
| 收稿时间： | 2021/5/22 0:00:00 |
| 修稿时间： | 2021/5/22 0:00:00 |
Study on tea origin traceability based on supervised and semi supervised models |
| |
| Authors: | HONG Xuezheng |
| |
| Abstract: | In order to protect the rights and interests of consumers and promote the development of tea industry, it is particularly important to identify the source of tea quickly and efficiently. At present, most of the tea traceability methods used in the market need a fixed number of samples for prediction and analysis, but in practical application, due to the demand of large-scale samples, the prediction cost will be further increased. This paper takes Zhejiang Anji white tea as the research object, through the comparison of the prediction accuracy of different sample sizes under different supervised and semi supervised models, in order to select the reasonable and economic sample size under the premise of ensuring the prediction accuracy in practical application. The results show that the tea traceability model based on semi supervised algorithm is better than that based on supervised algorithm when there are few labeled samples. |
| |
| Keywords: | Anji white tea Near infrared spectroscopy Origin traceability Sample size selection |
|
| 点击此处可从《现代农业科技》浏览原始摘要信息 |
| 点击此处可从《现代农业科技》下载免费的PDF全文 |
|
