Questions

1. What is the true probability distribution of an event? How can we discover what this distribution is?
2. Why is it more useful to conduct 1,000 trials of an experiment than 10 trials? Give a general explanation in 2–3 sentences.
3. Explain the importance of having a large enough sample size when conducting an experiment.
4. Explain what Thompson sampling is in 2–3 sentences.
5. Explain how the Bayesian control rule relates to Thompson sampling.
6. Briefly explain the difference between A/B testing and MAB testing.
7. Why are user ad-clicks a useful demonstration of this type of testing methodology? What might another good example problem be?