A car manufacturer uses an AI model to predict maintenance needs for its vehicles. Which of the following techniques can an IS auditor apply to MOST effectively verify the AI model's decisions to stak

Question

Options

• AUsing neural network visualization to show how the AI model processes data through its layers
• BUsing K-means algorithms to group vehicles based on mileage or engine temperature for
• CUtilizing support vector machines (SVM) to classify vehicles based on maintenance urgency
• DUsing local interpretable model-agnostic explanation (LIME) to analyze how specific features

• A
  5% (1)
• B
  5% (1)
• C
  10% (2)
• D
  80% (16)

Explanation

LIME (Local Interpretable Model-Agnostic Explanations) generates human-readable, feature-level explanations for individual predictions by approximating the complex model locally with a simpler interpretable model. It can explain why a specific vehicle was flagged for maintenance-identifying which features (e.g., mileage, temperature, vibration) most influenced the prediction-making it accessible to non-technical stakeholders. Neural network visualization (A) reveals internal layer activations, which are opaque to business stakeholders. K-means (B) is a clustering algorithm, not an explanation method. SVM (C) classifies but does not inherently provide the transparent, instance-level explanations needed for stakeholder communication.

No community discussion yet for this question.