Objective 1.2 Analyze the Use Cases of ML Algorithms to Rank Them by Their Success Probability AIP-210 Practice Questions (Page 4)

Questions 16–20

1. 16foundation · medium

   What does 'data readiness' refer to when estimating the success probability of an ML algorithm?

   • AThe data is stored in a cloud database
   • BThe data is accessible to all employees
   • CThe data is clean, labeled, and sufficient for training
   • DThe data is backed up daily
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A financial firm wants to segment its customers into distinct groups for targeted marketing. They have demographic and transaction data for 500,000 customers. They want to understand the characteristics of each segment. Which algorithm should be ranked highest for success probability?

   • ANaive Bayes classification
   • BK-means clustering
   • CLinear regression
   • DApriori association rule mining
   Reveal Solution Discussion Select an answer first
4. 18expert · hard

   A logistics company wants to predict delivery delays. They have historical data on delivery routes, weather, traffic, and package details. The data is highly dimensional and includes both numerical and categorical features. The model will be used to proactively reroute packages. Which problem type is this, and which algorithm is most appropriate?

   • AClustering; use DBSCAN
   • BRegression; use gradient-boosted trees
   • CClassification; use logistic regression
   • DRegression; use linear regression
   Reveal Solution Discussion Select an answer first
5. 19expert · hard

   A marketing team wants to build a model to predict which customers will respond to a campaign. They have a dataset of 100,000 customers with 100 features. The response rate is 10%. The team needs to maximize the number of true positives while keeping the false positive rate below 20%. They have a limited budget for model development. Which algorithm should be ranked highest for success probability?

   • AA random forest with hyperparameter tuning
   • BA deep neural network with dropout
   • CA gradient-boosted tree model with cost-sensitive learning
   • DA logistic regression with a custom threshold
   Reveal Solution Discussion Select an answer first
6. 20foundation · medium

   Which factor is a potential risk that could lower the success probability of an ML algorithm?

   • AAdequate computational resources
   • BHigh-quality labeled data
   • CData leakage in the training set
   • DA well-defined business objective
   Reveal Solution Discussion Select an answer first