ML Quiz 9

What is one-hot encoding?

• A vector representation where a single value is 1 and the rest are 0

• A vector representation where one value is 0 and the rest are 1

• A vector representation where multiple values are 1

Sparse representation is used when the binary vector of a feature is very large and it stores only the nonzero values.

• True

• False

Qualities of good features:

• A feature value should appear at least 5 times in a data set

• A feature and it's value should be obvious only to the engineer who designed it

• Avoid 'magic' values mixed with real data

• The definition of a feature shouldn't change over time

Scaling feature benefits:

• Helps gradient descent coverage more quickly

• Helps avoid the "NaN trap"

• Helps the model learn appropriate weights for each feature.

• None of the above

In real-life, many examples in data sets are unreliable due to one or more of the following:

• Omitted values

• Duplicate examples

• Bad labels

• Bad feature values

• None of the above

A feature cross is a synthetic feature that encodes nonlinearity in the feature space by multiplying two or more input features together.

• True

• False

What kinds of feature crosses can we create?

• a feature cross formed by multiplying the values of two features

• a feature cross formed by multiplying the values of five features

• a feature cross formed by squaring a single feature

• None of these

Feature crosses of one-hot feature vectors are similar to logical conjunctions.

• True

• False

Check all the good features.

• house_type: victorian

• unique_house_id: 8SK982ZZ1242Z

• house_age_years: 27

• house_age: 851472000

• city_id: "br/sao_paulo"

Binning method is used to smoothing data or to handle noisy data.

• True

• False