A hybrid transfer learning algorithm incorporating TrSVM with GASEN

作者：

Highlights：

• TrSVM is a novel transfer learning model based on Support Vector Machine.

• TrSVM constructs a representative subset of source data by using support vectors.

• TrSVM discards redundant data and effectively reduces the computational cost.

• TrGASVM assigns corresponding weights to different support vectors in the representative subset of source data.

• TrGASVM accomplishes the pruning process by incorporating the principle of GASEN.

摘要

•TrSVM is a novel transfer learning model based on Support Vector Machine.•TrSVM constructs a representative subset of source data by using support vectors.•TrSVM discards redundant data and effectively reduces the computational cost.•TrGASVM assigns corresponding weights to different support vectors in the representative subset of source data.•TrGASVM accomplishes the pruning process by incorporating the principle of GASEN.

论文关键词：Transfer learning,Ensemble selection,Genetic Algorithm based Selective Ensemble algorithm (GASEN),Support Vector Machine based Transfer Learning algorithm (TrSVM),Transfer Learning algorithm incorporating TrSVM with GASEN (TrGASVM)

论文评审过程：Received 4 August 2017, Revised 3 March 2019, Accepted 24 March 2019, Available online 28 March 2019, Version of Record 5 April 2019.

论文官网地址：https://doi.org/10.1016/j.patcog.2019.03.027