Joining iSAX indexes
At this point, we have iSAX indexes that we want to use to perform basic time series data mining tasks. One of them is finding similar subsequences between two or more time series. In our case, we are working with two time series, but the method can be extended to more time series with small changes.
How to join iSAX indexes
Given two or more iSAX indexes, it is up to us to decide how and why we are going to join them. We can even join them using SAX representations with a cardinality value of 2. However, using the SAX representations of the nodes as our keys for the join is the most logical choice.In our case, we are going to use the iSAX indexes and the SAX representations of the nodes to look for similar subsequences. This is because we have the intuition that subsequences in nodes with the same SAX representation are close to each other. The term close is defined relative to a distance metric. For the purposes of this chapter, we are going to use the...