EasyDeepLearn

What is Dynamic Time Warping?

hard

Answer

  • Similarity between two sequences allowing non-linear alignment (stretch/compress).
  • Finds the best matching via dynamic programming through a cost matrix.
  • Robust to phase shifts and different lengths.
  • Cost O(mn) per pair — expensive for large series (mitigated  by  lowerbounds  like  LBKeogh,  sakoechiba  constraint  window)(\mathrm{mitigated}\;\mathrm{by}\;\mathrm{lower} - \mathrm{bounds}\;\mathrm{like}\;\mathrm{LB}_{\mathrm{Keogh}}, \;\mathrm{sakoe} - \mathrm{chiba}\;\mathrm{constraint}\;\mathrm{window}).
  • Standard for shape-based time-series similarity in speech, ECG, gesture recognition.
Check yourself — multiple choice
  • Same as Euclidean
  • DP alignment allowing stretch/compress → phase-shift-robust sequence distance; O(mn); LBKeogh  /  warping\mathrm{LB}_{\mathrm{Keogh}}\; / \;\mathrm{warping} window for scale
  • Random
  • Not real

DTW: dynamic-programming alignment; robust to phase shifts.

#clustering#similarity

Practise Unsupervised Learning

214 interview questions in this topic.

Related questions