Feature: Dynamic Time Warping for sorting and clustering

[skim0119]

Feature

• Add dynamic time warping distance calculation function
  • After waveform extraction, a distance matrix can be computed using miv.signal.similarity.dynamic_time_warping_distance_matrix function.
  • The distance matrix can be passed to any sklearn unsupervised clustering algorithm.
• Add dtaidistance dependency in optional sortingExtension extension.
• Documentation update regarding DTW