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 1910.10682cs.LGcs.SIstat.ML
收藏

图神经网络的特征选择与特征提取

Feature Selection and Extraction for Graph Neural Networks

University of Alabama in Huntsville(阿拉巴马大学亨茨维尔分校)

Deepak Bhaskar Acharya, Huaming Zhang

AI总结:

本文将基于Gumbel Softmax的特征选择算法扩展到图神经网络,并提出特征排序机制,在Cora、Citeseer和Pubmed数据集上验证了特征选择与排序的有效性。

摘要:

图神经网络(GNNs)已成为数据科学领域最新的研究热点,原因在于它们以普遍存在的数据结构——图——作为构建和训练神经网络的底层元素。在一个GNN中,每个节点都关联着大量特征。整个任务(例如分类或聚类)利用节点的特征在节点级或图级做出决策。在本文中,(1)我们将通过Gumbel Softmax提出的特征选择算法扩展到GNNs。我们使用多个基准数据集Cora、Citeseer和Pubmed,对我们的特征选择算法进行了一系列实验。(2)我们实现了一种对提取特征进行排序的机制。我们证明了我们的算法在特征选择和特征排序方面的有效性。对于Cora数据集,(1)我们使用该算法从1433个特征中选出225个特征。我们的实验结果表明了它们在相同分类问题上的有效性。(2)我们提取的特征是原始特征的线性组合,其中每个提取特征的系数均为非负且总和为一。我们提出了一种对提取特征进行排序的算法,其意义在于:当将这些特征用于相同的分类问题时,对于排名1-50、51-100、100-150和151-200范围内的提取特征,准确率会逐渐下降。

英文摘要:

Graph Neural Networks (GNNs) have been a latest hot research topic in data science, due to the fact that they use the ubiquitous data structure graphs as the underlying elements for constructing and training neural networks. In a GNN, each node has numerous features associated with it. The entire task (for example, classification, or clustering) utilizes the features of the nodes to make decisions, at node level or graph level. In this paper, (1) we extend the feature selection algorithm presented in via Gumbel Softmax to GNNs. We conduct a series of experiments on our feature selection algorithms, using various benchmark datasets: Cora, Citeseer and Pubmed. (2) We implement a mechanism to rank the extracted features. We demonstrate the effectiveness of our algorithms, for both feature selection and ranking. For the Cora dataset, (1) we use the algorithm to select 225 features out of 1433 features. Our experimental results demonstrate their effectiveness for the same classification problem. (2) We extract features such that they are linear combinations of the original features, where the coefficients for each extracted features are non-negative and sum up to one. We propose an algorithm to rank the extracted features in the sense that when using them for the same classification problem, the accuracy goes down gradually for the extracted features within the rank 1 - 50, 51 - 100, 100 - 150, and 151 - 200.

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。

↑