How Data, Algorithms, and Models Form AI Foundations AIE Practice Questions (Page 7)

Questions 31–35

1. 31expert · hard

   A telecom company wants to predict customer churn. They have a dataset with 1 million records but only 5% are churners. They plan to train a model. What is the most important data-related issue to address?

   • AThe dataset is too large for any algorithm to process
   • BThe algorithm will overfit because there are too many records
   • CThe data is too clean and needs more noise
   • DThe class imbalance may cause the model to predict the majority class and ignore churners
   Reveal Solution Discussion Select an answer first
2. 32application · medium

   A bank wants to use AI to approve loan applications. They have historical data on approved and rejected loans. They decide to use a linear regression algorithm. What is the main issue with this choice?

   • ALinear regression requires a very large dataset to work
   • BLinear regression is designed for continuous outputs, not binary classification
   • CLinear regression will overfit the training data
   • DLinear regression cannot process numerical data
   Reveal Solution Discussion Select an answer first
4. 33expert · hard

   A robotics company wants to teach a robot to navigate a warehouse. They have a simulation environment that generates training data, but the real warehouse has different lighting and obstacles. The model performs well in simulation but poorly in the real world. What is the most effective strategy?

   • ADisable the model and use manual control
   • BCollect more simulation data to improve the model
   • CUse a more complex algorithm in simulation
   • DUse domain adaptation techniques to make the model generalize from simulation to real-world data
   Reveal Solution Discussion Select an answer first
5. 34foundation · easy

   Which sequence correctly represents the relationship among data, algorithms, and models in AI?

   • AAlgorithm → Data → Model
   • BData → Model → Algorithm
   • CModel → Algorithm → Data
   • DData → Algorithm → Model
   Reveal Solution Discussion Select an answer first
6. 35application · medium

   A bank wants to detect fraudulent transactions. They have a large dataset of historical transactions labeled as fraudulent or legitimate. They use a decision tree algorithm to create a model that classifies new transactions. Which statement best describes how data, algorithm, and model work together in this system?

   • AThe model generates new data that the algorithm uses to improve itself
   • BThe algorithm directly classifies new transactions without using the model
   • CThe data is used to validate the model, but the algorithm is independent of the data
   • DThe algorithm learns patterns from the historical data and encodes them into the model, which then classifies new transactions
   Reveal Solution Discussion Select an answer first