A scheme for high level data classification using random walk and network measures

作者：

Highlights：

• A new network-based classification technique is proposed.

• It considers the structural pattern formation of each data class.

• Physical data features and structural information are combined to characterize data.

• This mixed information improved classification accuracy.

摘要

•A new network-based classification technique is proposed.•It considers the structural pattern formation of each data class.•Physical data features and structural information are combined to characterize data.•This mixed information improved classification accuracy.

论文关键词：Supervised learning,Data classification,Network-based learning,High level classification,Markov chain,Random walk,Limiting probabilities,Steady states

论文评审过程：Received 28 August 2015, Revised 17 August 2017, Accepted 9 September 2017, Available online 15 September 2017, Version of Record 2 October 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.09.014