GAPORE: Boolean network inference using a genetic algorithm with novel polynomial representation and encoding scheme

作者：

Highlights：

• A novel genetic algorithm incorporating local search is introduced to infer Boolean networks accurately.

• An efficient symbolic polynomial representation is proposed to represent the unknown Boolean functions.

• A novel encoding scheme is developed to encode the Boolean functions flexibly.

• An l2-norm regularization is designed to reduce the over-fit problem in Boolean network inference.

摘要

•A novel genetic algorithm incorporating local search is introduced to infer Boolean networks accurately.•An efficient symbolic polynomial representation is proposed to represent the unknown Boolean functions.•A novel encoding scheme is developed to encode the Boolean functions flexibly.•An l2-norm regularization is designed to reduce the over-fit problem in Boolean network inference.

论文关键词：Data-driven,Boolean network,Polynomial Boolean representation,Hybrid genetic algorithm,Dominant bit encoding scheme

论文评审过程：Received 9 May 2021, Revised 22 June 2021, Accepted 30 June 2021, Available online 3 July 2021, Version of Record 9 July 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.107277