Adaptive feature selection based on the most informative graph-based features


In this paper, we propose a novel method to adaptively select the most informative and least redundant feature subset, which has strong discriminating power with respect to the target label. Unlike most traditional methods using vectorial features, our proposed approach is based on graph-based features and thus incorporates the relationships between feature samples into the feature selection process. To efficiently encapsulate the main characteristics of the graph-based features, we probe each graph structure using the steady state random walk and compute a probability distribution of the walk visiting the vertices. Furthermore, we propose a new information theoretic criterion to measure the joint relevance of different pairwise feature combinations with respect to the target feature, through the Jensen-Shannon divergence measure between the probability distributions from the random walk on different graphs. By solving a quadratic programming problem, we use the new measure to automatically locate the subset of the most informative features, that have both low redundancy and strong discriminating power. Unlike most existing state-of-the-art feature selection methods, the proposed information theoretic feature selection method can accommodate both continuous and discrete target features. Experiments on the problem of P2P lending platforms in China demonstrate the effectiveness of the proposed method.

Publication DOI:
Divisions: College of Engineering & Physical Sciences
?? 50811700Jl ??
College of Engineering & Physical Sciences > Systems analytics research institute (SARI)
Event Title: 11th IAPR-TC-15 International Workshop on Graph-Based Representations in Pattern Recognition, GbRPR 2017
Event Type: Other
Event Dates: 2017-05-16 - 2017-05-18
Uncontrolled Keywords: Theoretical Computer Science,Computer Science(all)
ISBN: 978-3-319-58960-2, 978-3-319-58961-9
Last Modified: 11 Dec 2023 08:41
Date Deposited: 13 Jun 2017 08:55
Full Text Link:
Related URLs: http://www.scop ... tnerID=8YFLogxK (Scopus URL)
PURE Output Type: Conference contribution
Published Date: 2017
Published Online Date: 2017-05-10
Accepted Date: 2017-03-06
Authors: Bai, Lu
Cui, Lixin
Rossi, Luca (ORCID Profile 0000-0002-6116-9761)
Hancock, Edwin R.
Jiao, Yuhang



Version: Accepted Version

| Preview

Export / Share Citation


Additional statistics for this record