Q1 (2pts): Refer to the “Lunch features” dataset to give an example of each data type:
Q1a: Numerical and Discrete:
Q1b: Numerical and Continuous:
Q1c: Categorical and Nominal:
Q1d: Categorical and Ordinal:
Q2 (5pts): You’ve been tasked with inputting the “Lunch features” dataset into a new database that can only accept numerical feature values. You must keep a minimum of 5 features in addition to price, but it’s fine to leave null values for samples that do not have a feature value recorded. List the features you’ll choose to keep and how you would process them for input:
Feature Processing
Price No processing necessary, just input the decimal value in dollar units
Q2a:
Q2b:
Q2c:
Q2d:
Q2e:
Q3 (2pts): Identify a data quality problem in the Spring subset of the “Lunch features” dataset. Propose a method to handle it.
Q4 (6pts): Within the “Lunch features” dataset, the Spring subset has many more features than the Fall subset. To integrate the two into a single matrix, you could either drop all extra features from the Spring samples or add all the features to the Fall samples. Answer three of the following with unique reasons:
Q4a: Why would dropping all extra features from the Spring samples would be a good idea?
Q4b: Why would dropping all extra features from the Spring samples would be a bad idea?
Q4c: Why would adding all extra features to the Fall samples would be a good idea?
Q4d: Why would adding all extra features to the Fall samples would be a bad idea?
