Which of the following approaches would help overcome testing challenges associated with probabilistic and non-deterministic AI-based systems?

Question

Options

• ARun the test several times to ensure that the AI always returns the same correct test result.
• BDecompose the system test into multiple data ingestion tests to determine if the AI system is
• CDecompose the system test into multiple data ingestion tests to determine if the AI system is
• DRun the test several times to generate a statistically valid test result to ensure that an appropriate

• A
  3% (2)
• B
  10% (6)
• C
  16% (10)
• D
  71% (44)

Explanation

Probabilistic and non-deterministic AI-based systems do not always produce the same output for identical inputs. This makes traditional testing approaches ineffective. Instead, the best approach is to run tests multiple times and analyze results statistically. Statistical Validity: Running tests multiple times ensures that observed results are statistically significant. Instead of relying on a single test run, analyzing multiple iterations helps determine trends, probabilities, and outliers. Expected Result Tolerance: AI-based systems may produce different results within an acceptable range. Defining acceptable tolerances (e.g., "result must be within 2% of the optimal value") improves test effectiveness.

No community discussion yet for this question.