Other Applications DY0-001 Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   What is a key challenge when implementing fraud detection algorithms?

   • AHandling missing values in the dataset
   • BHandling datasets with too many features
   • CHandling imbalanced datasets where fraud is rare
   • DHandling categorical variables only
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   Which industry commonly uses graph analysis to identify communities or clusters of related entities?

   • AHospitality
   • BAgriculture
   • CConstruction
   • DSocial media
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   A network security team wants to detect unusual patterns in network traffic that may indicate a cyber attack. They have a large dataset of normal network traffic and need to identify deviations from this baseline. Which anomaly detection method is most appropriate?

   • AK-nearest neighbors (KNN), a supervised classification algorithm that requires labeled data for both normal and anomalous traffic.
   • BOne-class SVM, which learns a boundary around the normal data and flags points outside it as anomalies.
   • CLinear regression, which models the relationship between network features and predicts normal traffic.
   • DApriori algorithm, which finds frequent itemsets in network traffic data.
   Reveal Solution Discussion Select an answer first
5. 14expert · hard

   A game development company is training an AI agent to play a complex strategy game. The game has a large state space and a continuous action space (e.g., movement speed and direction). The agent must learn a policy that maximizes the cumulative reward over a long time horizon. Which reinforcement learning algorithm is most appropriate?

   • AQ-learning with a tabular representation of the state-action space.
   • BSARSA, an on-policy temporal difference learning algorithm.
   • CProximal Policy Optimization (PPO), a policy gradient method that handles continuous action spaces and is stable for complex tasks.
   • DA rule-based system with predefined strategies for the game.
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   What is a common characteristic of fraudulent activities that data science methods exploit?

   • AThey follow the same patterns as normal activities
   • BThey occur with high frequency in most datasets
   • CThey are always easily identifiable by simple rules
   • DThey are rare and often deviate from normal patterns
   Reveal Solution Discussion Select an answer first