arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2205.05296cs.LG
收藏

子空间学习机(SLM):方法与性能

Subspace Learning Machine (SLM): Methodology and Performance

University of Southern California(南加州大学);Army Research Laboratory(陆军研究实验室)

Hongyu Fu, Yijing Yang, Vinod K. Mishra, C. -C. Jay Kuo

摘要:

受前馈多层感知器(FF-MLP)、决策树(DT)和极限学习机(ELM)的启发,本文提出了一种新的分类模型--子空间学习机(SLM)。SLM首先通过检查每个输入特征的判别能力来识别判别子空间$S^0$。然后,它使用$S^0$中的特征的概率投影来产生1D子空间,并为每个子空间找到最佳划分。这相当于用超平面划分$S^0$。提出了一个选择最佳$q$划分的准则,该准则能在它们之间产生$2q$划分子空间.我们将$S^0$分配给决策树的根结点,将$2q$子空间的交集分配给深度为1的子节点。分区过程递归地应用于每个子节点以构建SLM树。当子节点处的样本足够纯时,分区过程停止,并且每个叶节点进行预测。这个思想可以推广到回归,导致子空间学习回归器(SLR)。此外,SLM/SLR树的集合可以产生更强的预测器。SLM/SLR树,合奏和经典的分类器/回归器之间的性能基准进行了广泛的实验。

英文摘要:

Inspired by the feedforward multilayer perceptron (FF-MLP), decision tree (DT) and extreme learning machine (ELM), a new classification model, called the subspace learning machine (SLM), is proposed in this work. SLM first identifies a discriminant subspace, $S^0$, by examining the discriminant power of each input feature. Then, it uses probabilistic projections of features in $S^0$ to yield 1D subspaces and finds the optimal partition for each of them. This is equivalent to partitioning $S^0$ with hyperplanes. A criterion is developed to choose the best $q$ partitions that yield $2q$ partitioned subspaces among them. We assign $S^0$ to the root node of a decision tree and the intersections of $2q$ subspaces to its child nodes of depth one. The partitioning process is recursively applied at each child node to build an SLM tree. When the samples at a child node are sufficiently pure, the partitioning process stops and each leaf node makes a prediction. The idea can be generalized to regression, leading to the subspace learning regressor (SLR). Furthermore, ensembles of SLM/SLR trees can yield a stronger predictor. Extensive experiments are conducted for performance benchmarking among SLM/SLR trees, ensembles and classical classifiers/regressors.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑