Online Active Extreme Learning Machine With Discrepancy Sampling for PolSAR Classification
Online Active Extreme Learning Machine With Discrepancy Sampling for PolSAR Classification
复制标题
用于 PolSAR 分类的差异采样在线主动极限学习机
DOI：
10.1109/tgrs.2019.2952236
复制
发表时间：
2020-03
影响因子：
8.2
通讯作者：
Yang Shuyuan
中科院分区：
文献类型：
--
作者：
Li Lingling;Zeng Jie;Jiao Licheng;Liang Pujiang;Liu Fang;Yang Shuyuan
极限学习机（extreme learning machine，ELM）以其高精度和高效的学习能力，在机器学习领域受到越来越多的关注。然而，经典的ELM工作在批处理和被动的学习范式，它不能有效地处理顺序数据。ELM已经扩展到在线顺序学习形式（OS-ELM）和主动学习形式（AL-ELM），其中前者用于提高训练效率，后者主要用于提高精度。针对极化合成孔径雷达（PolSAR）图像分类中样本标注困难、代价高、样本有效性差、需要不断迭代学习等问题，提出一种在线主动极限学习机（OA-ELM）算法，联合收割机结合OS-ELM和AL-ELM的优点，弥补二者的不足，提高了算法的效率和泛化能力。OA-ELM能够动态地从序列数据中学习，具有较低的计算复杂度和较好的泛化能力。具体来说，OA-ELM通过扩展的递归最小二乘优化来减少训练的时间和内存成本。它还使用建议的差异采样（DS），修改了主动查询方法称为边缘采样（MS）选择的信息训练样本，提高了准确性。在将MS应用于ELM之前，需要首先将ELM的实值输出转换为概率输出。相反，建议的DS可以直接应用于ELM通过计算ELM的两个最大的实际非概率输出之间的差异。PolSAR分类实验结果表明，OA-ELM算法在分类精度和运行时间方面均优于其他算法。
The extreme learning machine (ELM) has drawn increasing attention in the field of machine learning due to its high accuracy and efficient learning. However, classical ELM works in batch and passive learning paradigms, which cannot deal with sequential data effectively. ELM has been extended to online sequential learning form (OS-ELM) and active learning form (AL-ELM), in which the former is used to improve training efficiency and the latter is mainly adopted to improve accuracy. In order to solve the problem of labeling samples difficulty and costly, poor sample validity, and continuous iterative learning in polarimetric synthetic aperture radar (PolSAR) image classification, we propose an online active extreme learning machine (OA-ELM) algorithm to combine the strengths and make up the weaknesses of OS-ELM and AL-ELM, which improves both efficiency and generalization ability. OA-ELM can learn from sequential data dynamically with low computational complexity and good generalization ability. Specifically, OA-ELM reduces time and memory cost for training via extended recursive least squares for optimization. It also improves accuracy using informative training samples selected by proposed discrepancy sampling (DS), which modifies an active query method called margin sampling (MS). Before applying MS to ELM, real-valued outputs of ELM need to be converted into probabilistic outputs first. Instead, the proposed DS can be applied to ELM directly by calculating the difference between the two largest actual nonprobabilistic outputs of ELM. Experimental results of PolSAR classification demonstrate that OA-ELM is effective and efficient compared with other algorithms in terms of accuracy and running time.
登录
查看更多内容
影响因子：
--
作者：
Liang, Nan-Ying;Huang, Guang-Bin;Sundararajan, N.
通讯作者：
Sundararajan, N.
影响因子：
8.2
作者：
A. Masjedi;M. J. V. Zoej;Y. Maghsoudi
通讯作者：
A. Masjedi;M. J. V. Zoej;Y. Maghsoudi
DOI：
10.1109/tnnls.2015.2424995
发表时间：
2016-04-01
影响因子：
10.4
作者：
Tang, Jiexiong;Deng, Chenwei;Huang, Guang-Bin
通讯作者：
Huang, Guang-Bin
DOI：
10.1109/tgrs.2013.2258675
发表时间：
2014-04-01
影响因子：
8.2
作者：
Uhlmann, Stefan;Kiranyaz, Serkan
通讯作者：
Kiranyaz, Serkan
影响因子：
6
作者：
Deng, Wan-Yu;Ong, Yew-Soon;Zheng, Qing-Hua
通讯作者：
Zheng, Qing-Hua