IBM A1000-075: Foundations of AI Intermediate Practice Exam: Medium Difficulty 2025

A healthcare organization wants to implement an AI system that can interpret medical images to assist radiologists in detecting abnormalities. The system must explain which features in the image led to its conclusions. Which type of AI approach is MOST critical for this use case?

A retail company is using IBM Watson Discovery to analyze customer feedback from multiple sources. They notice that the system is returning irrelevant results for queries about 'apple products' because it's confusing fruit references with technology products. What feature should they configure to improve accuracy?

A data science team is building a classification model and observes that their training accuracy is 98% but their validation accuracy is only 72%. What problem does this indicate, and what is the MOST appropriate solution?

An insurance company is developing an AI model to automate claim approval decisions. During testing, they discover that the model approves claims at significantly different rates based on the applicant's zip code, which correlates with demographic factors. What AI governance principle is being violated, and what should be the PRIMARY remediation approach?

A company wants to use IBM Watson Assistant to create a virtual agent for customer support. The agent needs to handle both structured queries (order status, account balance) and unstructured conversations (product recommendations). Which combination of capabilities should they implement?

A machine learning engineer is selecting an algorithm for a problem where the goal is to group customers into segments without predefined categories, based on purchasing behavior patterns. Which type of learning approach and algorithm family is MOST appropriate?

An AI system trained on historical hiring data is being used to screen job applications. The development team discovers that the training data spans 20 years and reflects past hiring practices that were less inclusive. What is the MOST important step to ensure responsible AI deployment?

A financial services firm is implementing IBM Watson to analyze market sentiment from news articles and social media. They need to process both English and Spanish content with equal accuracy. What consideration is MOST important for successful implementation?

A data scientist notices that their regression model performs well on average but makes significant errors on edge cases representing 5% of the data. The business considers these edge cases critical. What evaluation approach should they prioritize?

An organization is deploying an AI system that makes automated decisions affecting employee performance reviews. According to AI governance best practices, what framework component is MOST essential for this high-stakes application?