Objective 2.2 Explain Data Collection/transformation Process in ML Workflow (transformations Include Standardization; Normalization; Log, Square-Root, and Logit Transformations) AIP-210 Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   Why is it important to transform features that have very different scales in a dataset?

   • ATo ensure all features have the same mean and standard deviation
   • BTo convert categorical features into numerical ones
   • CTo make the dataset smaller and easier to store
   • DTo prevent features with larger values from dominating the model's learning process
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   A feature represents probabilities of an event, with values between 0 and 1. Which transformation is most appropriate to use as input to a linear model?

   • AStandardization
   • BLogit transformation
   • CLog transformation
   • DSquare-root transformation
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   In which machine learning model is the logit transformation most commonly used?

   • ALogistic regression
   • BDecision tree
   • CK-means clustering
   • DLinear regression
   Reveal Solution Discussion Select an answer first
5. 4expert · hard

   A data scientist is building a k-nearest neighbors (kNN) model with a dataset that contains outliers in the 'income' feature. The scientist is deciding between min-max normalization and standardization. Which consideration is most important when choosing between these two transformations for this model?

   • AMin-max normalization is more robust to outliers because it uses the minimum and maximum values.
   • BStandardization is preferred when the data has outliers because it is less sensitive to extreme values than min-max normalization.
   • CStandardization is more robust to outliers because it uses the mean and standard deviation, which are less affected by extreme values.
   • DMin-max normalization is preferred when the data has outliers because it compresses the range to [0,1].
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   A data scientist is working with a feature 'number_of_products_purchased' that is a count variable with a Poisson-like distribution (mean ≈ variance). The scientist wants to stabilize the variance and reduce skewness. Which transformation is most appropriate?

   • AStandardization (z-score)
   • BSquare-root transformation
   • CMin-max normalization
   • DLog transformation
   Reveal Solution Discussion Select an answer first