Hacker News Reader

Thanks for pointing. Yes it should be lower value to prevent over-fitting.

df_train_set.Income.value_counts() should be df_test_set probably in the part where it's comparing both.