摘要
由机器人与机器学习每日新闻的新闻记者兼工作人员新闻编辑-调查人员讨论人工智能的新发现。根据NewsRx记者来自韩国Geoje的新闻，研究表明："这项研究涉及一种使用机器学习方法和过采样技术对海水温度数据进行正常检测的方法。"这项研究的财政支持者包括韩国海洋和渔业部；Kiost项目；韩国政府；人工智能聚合公司Ovation人力资源开发。我们的新闻记者从韩国海洋科学技术研究所的研究中获得了一句话：“数据是在2017年至2023年期间使用太平洋、印度洋的碳导率-温度-深度(CTD)系统获得的。”海温资料由1414条剖面组成,其中G1218条正常剖面和196条异常剖面,该数据集存在异常数据量与正常剖面相比不足的不平衡问题,为此,我们采用了重复、均匀随机变量、合成少数过采样技术(SMOTE),利用过采样技术生成异常数据。和自动编码器(AE)的数据类平衡技术，以及基于Traine D四分位数区间(IQR)的单类支持向量机(OCSVM)，以及具有用于异常检测的平衡数据集的多层感知器(MLP)模型。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Investigators discuss new findings in artificial intelligence. According to news originating from Geoje, South Korea, by NewsRx correspondents, research stated, “This study deals with a method for a nomaly detection in seawater temperature data using machine learning methods wit h oversampling techniques.” Financial supporters for this research include Ministry of Oceans And Fisheries Korea; Kiost Projects; Korea Government; Artificial Intelligence Convergence Inn ovation Human Resources Development. Our news correspondents obtained a quote from the research from Korea Institute of Ocean Science and Technology: “Data were acquired from 2017 to 2023 using a C onductivity-Temperature-Depth (CTD) system in the Pacific Ocean, Indian Ocean, a nd Sea of Korea. The seawater temperature data consist of 1414 profiles includin g 1218 normal and 196 abnormal profiles. This dataset has an imbalance problem i n which the amount of abnormal data is insufficient compared to that of normal d ata. Therefore, we generated abnormal data with oversampling techniques using du plication, uniform random variable, Synthetic Minority Oversampling Technique (S MOTE), and autoencoder (AE) techniques for the balance of data class, and traine d Interquartile Range (IQR)-based, one-class support vector machine (OCSVM), and Multi-Layer Perceptron (MLP) models with a balanced dataset for anomaly detecti on.”
NETL
NSTL