We're going to start this lesson with an overview of the business cas you're going to be modeling with your investor classifier. Here's the situation you are an investment banking analyst on a team that executes multi-billion dollar debt capital markets transactions for large corporate clients. Your particular debt product, the syndicated revolver, depends on an interesting dynamic between issuers who are your large corporate clients and investors who are global financial institutions. To illustrate, imagine that your client is Verizon. Verizon pays millions of dollars each year in fees to global financial institutions for various investment banking services, such as merger and acquisition advisory fees, asset divestiture advisory fees, and debt, and equity capital markets underwriting fees. Verizon's multimillion dollar annual budget for investment banking fees is referred to as their fee wallet, and every investment bank on Wall Street is fighting for a piece of it. Verizon knows that their massive fee wallet gives them leverage over Wall Street firms, so they use it to demand cheap debt from these banks in the form of a syndicated revolver. The interest rate on a syndicated revolver is often below the bank's cost of capital, meaning that each financial institution is actually losing money by committing capital to a syndicated revolver if they're losing money by committing. Why would banks ever agree to this? The deal is if a bank like JP Morgan or Barclays wants to pitch Verizon for business, for example, to lead their next $1.5 billion bond issuance for a seven figure fee, the bank must commit to provide a portion of Verizon syndicated revolver bankers for this product call it pay-to-play. If JP Morgan refuses to participate in the syndicated revolver, but Barclays commits to provide a portion, Barclays will then have an opportunity to pitch their services to Verizon while the door shuts for JP Morgan. Typically, the amount of other investment banking business awarded to each firm is proportionate to the size of their commitment to the revolver, so this is a very important strategic product. Since your team arranges and executes syndicated revolver transactions, it's your responsibility to decide which investors to invite to participate, determine how much capital you should invite them to commit and anticipate their responses. In addition, your team is responsible for advising your client on the pricing and structure of that client's syndicated revolver. Essentially, what you're doing is brokering access to your client and determining the price. Each investment bank must pay-to-play by providing cheap debt capital. It's critical that you accurately predict investor behavior because if too many banks decline to participate, you won't raise enough capital for your client resulting in a failed transaction. On the other hand, if more banks commit than you expect, your client may end up with too many mouths to feed and not enough fee wallet of other investment banking business to go around. On the other hand, if you can accurately predict investor behavior, you can confidently invite the right investors at the right amounts and fine tune pricing and structure to ensure the success of the transaction for your client. The good news is that your team has years of historical data for thousands of transactions, including records of how each investment bank reacted when presented with an invitation to commit capital to a new syndicated revolver. Now, it's your job to use this data to construct a machine learning classification model to predict investor behavior. In this project, you're going to spend a little more time on data exploration to get a better sense for the relationships you'll be modeling. You'll also learn some new ways to slice and dice and visualize your data to make these relationships easier to see and understand. In the files for this lesson, you should have received investor data dot csv. This file contains transaction data for five of your most active investor. Each observation represents one investor's reaction to an invitation to participate in a transaction. You can think of each row of data as one investor response.