Objective 1.2 Analyze the Use Cases of ML Algorithms to Rank Them by Their Success Probability AIP-210 Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A manufacturing company wants to predict equipment failure (maintenance) using sensor data. They have 1,000 labeled failure events and 100,000 normal operation records. The data is noisy and has missing values. Which factor should be considered MOST critical when ranking algorithms for this use case?

   • AThe interpretability of the model for engineers
   • BThe ability to deploy on edge devices
   • CThe speed of training on the available hardware
   • DThe algorithm's ability to handle imbalanced data and missing values
   Reveal Solution Discussion Select an answer first
2. 12expert · hard

   A credit card company wants to build a fraud detection model. They have a highly imbalanced dataset (0.1% fraud). The model must have a very low false positive rate to avoid annoying customers, but also catch as many fraud cases as possible. The deployment environment is a real-time transaction processing system with strict latency limits. Which algorithm should be ranked highest for success probability?

   • AA deep neural network with anomaly detection
   • BA random forest with balanced class weights
   • CA logistic regression with threshold tuning
   • DAn ensemble of gradient-boosted trees with cost-sensitive learning
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   A medical diagnostics company is developing a model to detect a rare disease from lab test results. The dataset has 5,000 positive cases and 95,000 negative cases. The cost of a false negative is extremely high (missed diagnosis). Which evaluation metric should be prioritized when comparing algorithms?

   • ARecall (sensitivity)
   • BPrecision
   • CAccuracy
   • DF1-score
   Reveal Solution Discussion Select an answer first
5. 14foundation · medium

   Which factor is LEAST likely to be considered when ranking ML algorithms for a use case?

   • AComputational cost
   • BRobustness to noise
   • CThe color of the algorithm's logo
   • DInterpretability
   Reveal Solution Discussion Select an answer first
6. 15foundation · medium

   When ranking ML algorithms for a use case, which criterion is most relevant for a healthcare diagnosis model where false negatives are dangerous?

   • AInterpretability
   • BTraining time
   • CRecall (sensitivity)
   • DModel size
   Reveal Solution Discussion Select an answer first