Why would they do that and where do you see evidence they did?
Because it's a direct way to measure predictive power, and it says so: "We’ll use leave-one-out cross-validation"
Because it's a direct way to measure predictive power, and it says so: "We’ll use leave-one-out cross-validation"