Objective 4.2 Train Models CERTIFIED-DATA-SCIENCE-PRACTITIONER Practice Questions (Page 5)

Questions 21–25

1. 21application · easy

   A machine learning engineer is tuning a random forest classifier and wants to explore a range of values for the number of trees and the maximum depth. They want to systematically evaluate all combinations. Which approach should they use?

   • AGrid search
   • BBayesian optimization
   • CManual tuning
   • DRandom search
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   What does the 'k' in k-means clustering represent?

   • AThe number of features in the dataset
   • BThe number of data points in the dataset
   • CThe number of iterations to run
   • DThe number of clusters to form
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   What is a hyperparameter in machine learning?

   • AA parameter learned from the training data
   • BA metric used to evaluate model performance
   • CA configuration setting set before training begins
   • DA feature that is derived from the raw data
   Reveal Solution Discussion Select an answer first
5. 24application · medium

   A logistics company wants to forecast weekly demand for their delivery services. They have historical data that shows a strong weekly seasonal pattern. Which model is most appropriate for this forecasting task?

   • AK-means clustering
   • BNaïve Bayes
   • CARIMA with seasonal components
   • DLinear regression
   Reveal Solution Discussion Select an answer first
6. 25expert · hard

   A geographic information system (GIS) analyst wants to identify clusters of earthquake epicenters. The data contains noise and clusters of varying shapes and densities. Which clustering algorithm is most appropriate?

   • ALinear regression
   • BK-means
   • CDBSCAN
   • DHierarchical clustering
   Reveal Solution Discussion Select an answer first