Fundamentals of testing AI-based systems · Question 26 of 60
With which statistical means can a model's accuracy, fairness and robustness be quantified?
3 answers are correct. Medium
Show answer
Answer
- A Error analysis
- B Hypothesis testing
- D Confidence intervals
Explanation
AI systems easily pick up biases from their data and can appear very sure of themselves although they are wrong. With statistical means such as confidence intervals, hypothesis tests and error analyses, accuracy, fairness and robustness can nevertheless be expressed in numbers.