1,510 | 53 | 141 |
下载次数 | 被引频次 | 阅读次数 |
One-hot编码是一种将类别值转换为易于在机器学习算法中合理计算类别值之间距离或相似度的编码的过程。介绍One-hot的编码规则,与标签编码进行比较,给出Python下One-hot编码的具体实现过程,指出One-hot编码的优缺点和One-hot编码维度过高的解决方法。
Abstract:One-hot encoding is a process of transforming category values into codes which are easy to calculate the distance or similarity between category values in machine learning algorithm. This paper introduces the encoding rules of One-hot encoding, compares it with label encoding, gives the specific implementation process of One-hot encoding based on Python, points out its advantages and disadvantages and solutions for reducing One-hot encoding dimensions.
1 Yuchen Qiao,Xu Yang,Enhong Wu.the research of BP Neural Network based on 0ne-Hot Encoding and Principle compOnent Analysis in determining the therapeutic effect of diabetes mellitus[J].IOP Conference Series:Earth and Environmental Science,2019,267(4):2-7.
2 Jia Li,Yujuan Si,Tao Xu,Saibiao Jiang,volodymyr Ponomaryov.Deep Convo1utional Neural Network Based ECG Classification System Using Information Fusion and One-hot Encoding Techniques[J].Mathematical Problems in Engineering,2018:1-10.
3 梁杰,陈嘉豪,张雪芹等.基于独热编码和卷积神经网络的异常检测[J].清华大学学报(自然科学版),2019,59(7):512-522.
4 孟杭,黄细霞,涂修建.基于时间序列和Xgboost的钢卷仓储吞吐量预测.计算机应用,2019,39(S2):24-28.
5 王济川,郭志刚.Logistic回归模型:方法与应用[M].北京:高等教育出版社,2001:2-11.
基本信息:
DOI:
中图分类号:TP181
引用信息:
[1]刘辉玲,陶洁,邱磊.基于Python的One-hot编码的实现[J].武汉船舶职业技术学院学报,2021,20(03):136-139.
基金信息:
武汉船舶职业技术学院院级科研项目“云计算在智慧校园建设中的应用”(项目编号:2017y08)