Summary

In this chapter, we introduced the dynamic time warping (DTW) algorithm, which is an excellent tool to find similarity between time series without any previous training. We presented an implementation of DTW to find similarity between a set of images, which worked very well in most cases. This method can be used for several other problems in a variety of areas such as robotics, computer vision, speech recognition, and time series analysis. We also saw how to turn an image into a time series with the PIL library. Finally we learned how to implement DTW with the mlpy library. In the next chapter, we will present how simulation can help us in the data analysis and how to model pseudo-random events.