Finally, we have worked through the entire machine learning algorithm, which can be summed up like this:

i$.$Randomly split the data into train and test sets

ii$.$Fit polynomial models of varying degrees on the train set

iii.$$Select the model that minimizes average error on the test set. Use this model to make predictions on new data.

iv$.$As you collect more data, automatically repeat the above steps $($to refine your model$)$

$$

$10)$ Using this algorithm, in what ways did the machine $$learn$$ the relationship between income and produce buying? In what ways is the model still imperfect? Explain.