RQ6 : Spatial deformations

RQ6.1

Are the methods robust to linear trends in the time series ?

In this section, we evaluate the robustness of methods to the presence of linear trends of different amplitudes. Our evaluation is summarized in the plot below.

Summary of the results

RQ6.1 Conclusion

SetFinder shows consistent performances in the presence of random walk. Motiflets, PEPA, and LoCoMotif maintain correct performance even for high random walk amplitudes.

RQ6.2

Are the methods robust to noise in the time series ?

In this section, we evaluate the robustness of methods to noise. Our evaluation is summarized in the plot below.

Summary of the results

RQ6.2 Conclusion

STOMP, PANMP, and Motiflets are the most suitable for very noisy time series. If the noise is non-zero without being excessively high, one can also consider using VALMOD or PEPA.