A segmental HMM based trajectory classification using genetic algorithm

作者：

Highlights：

• An improved multi-kernel using Convex Hull and Douglas Peucker algorithm is proposed.

• Classification is done with a two-stage HMM method using Global and Segmental HMM.

• The combination of two-stage HMM classification is done using a genetic algorithm.

• Experiments have been performed using two public datasets.

摘要

•An improved multi-kernel using Convex Hull and Douglas Peucker algorithm is proposed.•Classification is done with a two-stage HMM method using Global and Segmental HMM.•The combination of two-stage HMM classification is done using a genetic algorithm.•Experiments have been performed using two public datasets.

论文关键词：Trajectory classification,HMM,Segmental HMM,Supervised learning,Signature recognition,Genetic algorithm,SVC2004

论文评审过程：Received 30 June 2017, Revised 1 September 2017, Accepted 8 October 2017, Available online 12 October 2017, Version of Record 16 October 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.10.021