4.1 Explain AI’s Role in Exploratory Data Analysis (EDA) AI-TECHNICAL-PRACTITIONER Practice Questions (Page 4)

Questions 16–20

1. 16expert · hard

   A financial institution is building a credit risk model and has a dataset with hundreds of features, many of which are highly correlated. The team needs to reduce dimensionality while maintaining model interpretability for regulatory compliance. Which approach best balances these needs?

   • AApply principal component analysis (PCA) to create new composite features.
   • BUse feature importance from a tree-based model to select a subset of the most relevant features.
   • CKeep all features to avoid losing any information, even if it increases model complexity.
   • DUse an autoencoder to compress the features into a lower-dimensional representation.
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A data engineer receives a raw dataset from an external source and needs to assess its quality before it can be used in a machine learning pipeline. The engineer wants an automated way to understand data types, distributions, and potential issues. Which AI-driven approach is most appropriate?

   • AVisualize the data using a single histogram of the first column.
   • BManually inspect a random sample of rows to infer data quality.
   • CRun an AI-based data profiling tool that summarizes each column's distribution, detects anomalies, and flags missing values.
   • DImmediately train a model to see if the data can predict something, ignoring quality checks.
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   A data team is using an AI tool to automate EDA on a customer feedback dataset. The tool generates a summary of key themes and suggests that 'poor customer service' is the main driver of negative feedback. Before reporting this to management, what should the team do?

   • AAsk the AI tool to automatically send the report to management without review.
   • BIgnore the finding because AI cannot understand customer sentiment.
   • CReview the underlying feedback samples to confirm the theme is accurately represented.
   • DReport the finding as-is, since the AI tool is reliable.
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   Which of the following is an example of an AI-driven data profiling task?

   • AAutomatically identifying that a column contains date values instead of strings
   • BDeploying a machine learning model to production
   • CManually counting missing values in a spreadsheet
   • DWriting a report on business strategy
   Reveal Solution Discussion Select an answer first
6. 20foundation · easy

   How does AI support iterative hypothesis testing in EDA?

   • ABy automatically accepting or rejecting hypotheses without human review
   • BBy eliminating the need for data exploration
   • CBy creating hypotheses that are always correct
   • DBy suggesting hypotheses based on data patterns and then validating them against the data
   Reveal Solution Discussion Select an answer first