Discovery of Driving Behavior Patterns - Smart Information Systems: Computational Intelligence for Real-Life Applications - page 325

12.9.2 Synthetic Data
This controlled experiment aims at visualizing the clustering results of the proposed
RRR distance measure compared to the DTW distance.
We generated a labeled dataset, which consists of nine time series from three
different categories, called Wave, YoYo, and Peak. Each category comprises three
time series characterized by multiple occurrence of the same artificial patterns at
arbitrary positions. The dataset consists of univariate time series of equal length,
as shown in Fig. 12.3 . To visualize the clustering results of the RRR and DTW
(a)
( b)
Fig. 12.3 Univariate a and multivariate b synthetic time series with artificially implanted patterns
( red color ) at arbitrary positions, where each time series belongs to one of three groups (Wave,
YoYo , a n d P e a k )