The paper “Unsupervised Nonlinear Feature Selection from High-dimensional Signed Networks” published by Huang Qiang, a 2018 Master’s candidate under the supervision of Prof. Chang Yi, was accepted by the CCF-A conference (AAAI 2020) . AAAI Conference on Artificial Intelligence (AAAI) is one of the top academic conferences in the field of Artificial Intelligence.

This work was in collaboration with Prof. Makoto Yamada of Kyoto University in Japan.

Paper Information:

First Author: Huang Qiang

Title: Unsupervised Nonlinear Feature Selection from High-dimensional Signed Networks

Conference: 34^th AAAI Conference on Artificial Intelligence (AAAI 2020)

Conference Category: CCF-A

ConferenceDate : February 7-12, 2020, New York, USA

Summary:

In this paper, a new feature selection algorithm, SignedLasso, for symbolic networks is proposed. Using HSIC Lasso nonlinear feature selection operator, the features are mapped into high-dimensional space to capture the nonlinear dependencies between features and outputs. By combining the positive link and negative link information in the network, feature selection is carried out for the symbolic network.

                                                                                                       School of Artificial Intelligence

                                                                                                                 Nov 13, 2019