Draft AI Solution PMI-CPMAI Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   A business wants to segment its customers into distinct groups based on purchasing behavior, but there are no predefined labels. Which AI model type is most appropriate?

   • AReinforcement learning
   • BSupervised regression
   • CUnsupervised clustering
   • DSupervised classification
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   When defining data flow requirements, which processing approach is most appropriate for a scenario where data must be analyzed as it arrives to trigger immediate actions?

   • AOffline processing
   • BReal-time processing
   • CBatch processing
   • DMicro-batch processing
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   Which component of a high-level AI architecture is responsible for exposing the model's predictions to end users or other applications?

   • AModel serving layer
   • BData preprocessing layer
   • CModel training layer
   • DData storage layer
   Reveal Solution Discussion Select an answer first
5. 9expert · hard

   A multinational retailer wants to integrate an AI demand forecasting system with its existing ERP system. The ERP is hosted on-premises in each country, and the AI system will be in the cloud. Data privacy regulations require that personal data of customers remain in the country of origin. Which integration approach best meets these constraints?

   • ADeploy the AI system on-premises in each country, and use batch file transfers to a central cloud for model training.
   • BDeploy a single cloud AI system in one region, and use VPN connections to access all on-premises ERPs, transferring data to the central region.
   • CUse a hybrid integration platform that connects to all ERPs and stores data in a single cloud data lake, with encryption at rest.
   • DDeploy a cloud AI system with regional instances in each country, and use REST APIs to connect to each on-premises ERP, ensuring data stays within the region.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A retail company wants to build an AI system that predicts daily store footfall to optimize staff scheduling. They have historical sales data, weather data, and local event data. The solution must be deployed on-premises due to data privacy regulations. Which high-level architecture component is most critical to include in the initial design?

   • AA batch processing pipeline that ingests historical data, trains a supervised regression model, and serves predictions via a REST API to the scheduling system.
   • BA real-time streaming pipeline that ingests live footfall data, trains a reinforcement learning model, and outputs recommendations directly to staff mobile devices.
   • CAn edge computing solution that processes data on local IoT devices, uses unsupervised clustering to group days, and sends alerts when footfall deviates from normal.
   • DA cloud-based data lake that stores all raw data, uses a pre-trained generative model to simulate footfall, and provides a dashboard for manual analysis.
   Reveal Solution Discussion Select an answer first