Scholarlink

<p>A new procedure to automated error detection for Argo procedure is introduced to produce intermediate-quality dataset quickly. The procedure is one of the improve methods based on the previous study with a path-signature, in which we incorporate the new machine learning methods. The performances trained with the dataset produced in 2016 were similar to the ones based on 2022 dataset and robust from 2017 to 2021. This suggests that the method in this study was successfully learn the general features of QCs and can discriminate error profiles relatively close to the one in Argo data centers.</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

Improved Automated Quality Control Method based on the Signature for Argo profiles

  • 期刊论文

  • 发表日期：

    2026年4月29日

查看全文
  • Shinya Kouketsu,
  • Nozomi Sugiura,
  • Kanako Sato,
  • Shigeki Hosoda

摘要

A new procedure to automated error detection for Argo procedure is introduced to produce intermediate-quality dataset quickly. The procedure is one of the improve methods based on the previous study with a path-signature, in which we incorporate the new machine learning methods. The performances trained with the dataset produced in 2016 were similar to the ones based on 2022 dataset and robust from 2017 to 2021. This suggests that the method in this study was successfully learn the general features of QCs and can discriminate error profiles relatively close to the one in Argo data centers.