Time Series Indexing

5

Joining and Comparing iSAX Indexes

In the previous chapter, we developed a Python package called isax that creates iSAX indexes for indexing the subsequences of a time series given a sliding window.

In this chapter, we are going to experiment with how the sliding window size affects the number of splits and the number of accesses to subsequences while creating an iSAX index.

Then, we are going to use the iSAX indexes created by the isax package and try to join and compare them. By comparing, we aim to understand the efficiency of an iSAX index, and by joining, we mean being able to find similar nodes in two iSAX indexes based on SAX representations.

The last part of this chapter is going to briefly discuss Python testing before developing ...