An analysis of alignment and integral based kernels for machine learning from vessel trajectories

作者：

Highlights：

• Alignment kernels outperform integral over time kernels for vessel trajectories.

• Piecewise linear segmentation improves alignments, not integral over time.

• Softmax alignment kernels are good for outlier detection.

摘要

•Alignment kernels outperform integral over time kernels for vessel trajectories.•Piecewise linear segmentation improves alignments, not integral over time.•Softmax alignment kernels are good for outlier detection.

论文关键词：Alignment measures,Integral over time similarity,Vessel trajectories,Trajectory kernels

论文评审过程：Available online 1 June 2014.

论文官网地址：https://doi.org/10.1016/j.eswa.2014.05.025