引導式測驗 · 中等 等級

蒙特卡羅模擬測驗

測試隨機試驗估計、π 模擬、輸入分佈、相關性、樣本量縮放、罕見事件、再現性和限制。

相關引導路徑蒙特卡羅模擬
問題 1 的 8

Which sequence best matches the guide's Monte Carlo workflow?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. Which sequence best matches the guide's Monte Carlo workflow?
    • Replace all uncertain inputs with one fixed average value
    • Choose the most favorable possible outcome and report it as expected
    • Draw from a specified model repeatedly, compute each result and summarize them
    • Fit a model once and assume its output is an exact future event
  2. In the guide's unit-circle example, why is the fraction of sampled points inside the circle multiplied by four?
    • The unit circle has four times the area of its enclosing square
    • The fraction estimates π directly and the factor four corrects sample bias
    • The square's area is four while the unit circle's area is π
    • Each of four quadrants is estimated separately and then added
  3. An illustrative run places 785 of 1,000 points inside the unit circle. What estimate of π does it produce?
    • 0.785
    • 3.14
    • 4.00
    • 785.0
  4. Two project costs usually rise together. What error can result from sampling them as independent inputs?
    • Only run reproducibility changes; joint risk remains unaffected
    • Each input mean becomes biased but their joint extremes remain correct
    • Sampling the right marginal distributions automatically preserves co-movement
    • The simulated risk can misrepresent their real dependence
  5. For a simple independent sample mean with finite variance, roughly how many draws are needed to halve its standard error?
    • About four times as many
    • Twice as many
    • The same number with a new seed
    • One-quarter as many
  6. Why does not seeing a rare event in a small simulation fail to prove the event impossible?
    • Rare events require no input distribution
    • A simulation always excludes rare outcomes by definition
    • Recording a seed prevents rare events from occurring
    • Finite random draws may simply miss a low-probability outcome
  7. What does recording a pseudorandom seed allow a team to do?
    • Reproduce the same run while still testing robustness with other draws
    • Eliminate all sampling variation in future independent runs
    • Guarantee the simulated average equals the true quantity
    • Prove the input assumptions match reality
  8. A simulation returns a very narrow range after millions of draws. What limitation can remain?
    • A large run necessarily makes every input distribution correct
    • Incorrect input models or omitted dependencies can still bias the answer
    • The square-root rule stops applying to every quantity
    • The output must now be a prediction interval for one person