If your error on the test set is low, you have learned the weights and biases of a model that predict the output given specific inputs and you can now use it to predict unseen data. 9/10

→ View original post on X — @abacusai,