A hybrid approach to cardinality constraint portfolio selection problem based on nonlinear neural network and genetic algorithm

作者：

Highlights：

• In this work, Genetic Algorithm and Nonlinear Neural network are hybridized.

• Mixed Integer Quadratic Optimization problem is solved by the hybridized approach.

• The proposed hybrid approach overcomes the challenges of NP-hard problems.

• Information Ratio, Sharpe Ratio, Sortino Ratio, Average Return calculated.

• Istanbul Stock Exchange 30 datasets is used.

摘要

•In this work, Genetic Algorithm and Nonlinear Neural network are hybridized.•Mixed Integer Quadratic Optimization problem is solved by the hybridized approach.•The proposed hybrid approach overcomes the challenges of NP-hard problems.•Information Ratio, Sharpe Ratio, Sortino Ratio, Average Return calculated.•Istanbul Stock Exchange 30 datasets is used.

论文关键词：Portfolio optimization,Nonlinear neural network,ISE-30,Cardinality constraint

论文评审过程：Received 20 March 2020, Revised 7 December 2020, Accepted 17 December 2020, Available online 24 December 2020, Version of Record 16 January 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114517