Question

Suppose we trained a supervised learning algorithm on some training data and observed that the resultant model gave no error on the
training data. Which among the following conclusions can you draw in this scenario? (multiple options may be correct)
it is possible that the learned model will generalise well to unseen data
it is possible that the learned model will not generalise well to unseen data
the learned model will definitely perform well on unseen data
the learned model will definitely not perform well on unseen data

Answer 1

Explanation:

Among the various options given in question statement regarding supervised learning algorithm, below two options are the correct ones;

• it is possible that the learned model will generalise well to unseen data
• it is possible that the learned model will generalise well to unseen data

Both reasons are valid because if we have chosen a very simple training data then first condition is possible but if we have chosen an extremely complex training data then second condition will be possible