The original dataset only had two columns: address and balance. I’ve added five new ones. For address_tier, is_corporation, and is_retail, I applied a few logical rules.