R exercise

Naive Bayes 1

Sign in to hand in your solution and get feedback.

nb.fit <- naiveBayes(Direction ~ Lag1 + Lag2, data = Smarket, subset = train) nb.preds <- ...