Time-Weighted Dynamic Time Warping for Remote Sensing Time
Series Analysis
Description
Provides a Time-Weighted Dynamic Time Warping (TWDTW) algorithm to measure similarity between two temporal sequences. This adaptation of the classical Dynamic Time Warping (DTW) algorithm is flexible to compare events that have a strong time dependency, such as phenological stages of cropland systems and tropical forests. This package provides methods for visualization of minimum cost paths, time series alignment, and time intervals classification.