The stages of predictive analytics data

- [Narrator] So let's follow the data…through the whole process.…When the modeler begins, their almost always starting…by looking at lots of transactions.…So for instance, we have a bunch of items…that people have bought,…and we can see that several rows…are associated with one customer.…What they're looking for at this stage,…is things as simple as,…how long is the maximum length of time…a customers been around?…Did they buy just a few things,…or did they buy a lot of things?…Because, what we want to zero in on…is the individual customer level.…

So here we're looking at just one customer.…Mr. Streeter, who lives on Rockwell Lane,…and what we have to now do, is pull in…all the transactions that belong to just him.…So our data set is about to get shorter,…but much wider,…let's see why.…It's not just all those transactions…that we're interested in, but all kinds of other things…that we can get at that individual level.…So we might have had hundred of thousands of transactions,…but associated with ten's of thousands of customers.…