Check your understanding of this module before moving on.
?
0/3
Checkpoint quiz 3 questions
A quick check on this module's key ideas before you move on
Question 01
Why can engineering a single well-chosen feature, like a debt-to-income ratio, sometimes improve a model more than switching to a fancier algorithm?
B is correct. If the real signal is a ratio between two columns, a model given only the raw columns has to discover that relationship the hard way. Handing it the ratio directly is often a bigger lever than a more sophisticated algorithm.
Question 02
Why does one-hot encoding avoid assigning arbitrary integers like Red=1, Blue=2, Green=3 to categories?
B is correct. Assigning Green=3 implies it is "three times further" from Red than Blue is, which is meaningless for an unordered category. One-hot encoding avoids inventing that fake relationship.
Question 03
Why is median imputation often preferred over mean imputation for a skewed or outlier-prone column?
B is correct. The same robustness property from the statistics lesson applies here: one extreme value can drag a mean far from typical, but barely moves the median.