A multi-start algorithm to design a multi-class classifier for a multi-criteria ABC inventory classification problem

作者：

Highlights：

• The ABC classification problem is solved using a discrete artificial neural network.

• A randomized greedy multi-start algorithm is designed to train the neural network.

• The neurons’ weights are finding by solving linear programming formulations.

• The designed classifier shows good generalization accuracy for benchmark datasets.

• The proposed algorithm can be straight applied to other multi-class classification problems.

摘要

•The ABC classification problem is solved using a discrete artificial neural network.•A randomized greedy multi-start algorithm is designed to train the neural network.•The neurons’ weights are finding by solving linear programming formulations.•The designed classifier shows good generalization accuracy for benchmark datasets.•The proposed algorithm can be straight applied to other multi-class classification problems.

论文关键词：Multi-criteria ABC inventory classification,Multi-class classification,Neural networks,Multi-start algorithm

论文评审过程：Received 24 June 2016, Revised 1 October 2016, Accepted 18 February 2017, Available online 16 March 2017, Version of Record 26 March 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.02.048