Niching genetic network programming with rule accumulation for decision making: An evolutionary rule-based approach

作者:

Highlights:

• A novel learning classifier system named nGNP-RA is proposed.

• A niching mechanism is proposed to evolve the high-quality and diverse individuals.

• A reinforcement learning-based credit assignment method is developed.

• Experiments show the superiority of nGNP-RA.

• It provides a novel rule-based system for decision making tasks.

摘要

•A novel learning classifier system named nGNP-RA is proposed.•A niching mechanism is proposed to evolve the high-quality and diverse individuals.•A reinforcement learning-based credit assignment method is developed.•Experiments show the superiority of nGNP-RA.•It provides a novel rule-based system for decision making tasks.

论文关键词:Decision making,Genetic network programming,Learning classifier system,Niching,Reinforcement learning,Rule accumulation

论文评审过程:Received 2 May 2018, Revised 11 July 2018, Accepted 18 July 2018, Available online 26 July 2018, Version of Record 6 August 2018.

论文官网地址:https://doi.org/10.1016/j.eswa.2018.07.041