Let's take a look at the results. What does this tell you about the relationships in your data? Most of the time it looks like these five investors are invited at the bookrunner tier, which is the top tier of investors. That makes sense because these five are the most active investors in your product. When they're invited at the bookrunner tier, it looks like these banks rarely decline a transaction. When they're invited at the participant tier, it looks like they're much more likely to decline. Wouldn't knowing this relationship make you more likely to make better predictions about how these investors are going to behave when invited to a new transaction? Now, if I tell you an investor's invited to participate at the bookrunner tier, you have a lot of confidence that they're going to commit to the transaction. But if I tell you that they're being invited at the participant tier, you know that you need to be a little bit more cautious and not count on those dollars to come in to provide your issuer, your client with that capital. This preliminary process of exploring your data set to discover these types of relationships are what validate your hypotheses that this data actually does represent real relationships that you'll be able to pass to a machine learning algorithm, and that when your machine learning algorithm learns the relationships in this data, that it then will be able to make accurate predictions about how your investors will behave.