Bài kiểm tra có liên kết hướng dẫn · Trung bình Cấp độ

Bài kiểm tra ước tính độ co giãn giá của AI

Kiểm tra độ co giãn dưới dạng phản ứng nhu cầu, xác định nguyên nhân, tính nội sinh về giá, sự đánh đổi thử nghiệm, tính không đồng nhất và giới hạn của các ước tính chỉ dựa trên mô hình.

Đường dẫn hướng dẫn liên quanƯớc tính độ co giãn giá AI
Câu hỏi 1 của 8

What question does price elasticity estimate?

Xem trước mọi câu hỏi8 câu hỏi

Câu trả lời và giải thích xuất hiện khi bạn làm bài kiểm tra.

  1. What question does price elasticity estimate?
    • How advertising copy affects brand awareness only
    • How demand responds to a change in price
    • How product inventory changes when images are refreshed
    • How customer age predicts browsing duration
  2. Why can a regression of historical sales on price be biased for a pricing decision?
    • Prices are randomly assigned in all historical retail data
    • Sales quantities cannot be measured numerically
    • Machine learning always eliminates confounding
    • Sellers may set prices in response to expected demand or other changing factors
  3. What tradeoff does the 2023 ICML pricing-experiment paper study?
    • Learning price effects versus expected experiment revenue and tail risk
    • Choosing among product images without measuring demand
    • Maximizing training accuracy while ignoring the experiment outcome
    • Predicting customer identities from browsing data
  4. What can a randomized pricing experiment contribute?
    • A guaranteed revenue increase after any price change
    • Removal of the need to check legal or ethical constraints
    • Price variation that can help estimate a causal demand response when design assumptions hold
    • Proof that every customer responds identically
  5. How can machine learning help with elasticity estimates without solving everything?
    • It turns any historical correlation into a causal effect
    • It removes price endogeneity by adding more parameters
    • It guarantees a safe price for every product
    • It can model flexible patterns or heterogeneous responses, while causal identification still needs design and assumptions
  6. When should a pricing model avoid recommending a change?
    • When the team is monitoring a rollout
    • When available variation or data is too weak to support the estimate
    • When the model can produce a precise decimal
    • When products have different response patterns
  7. Why report uncertainty and segment results?
    • Uncertainty is irrelevant once a model has trained
    • A single estimate guarantees the same outcome everywhere
    • Average responses may hide meaningful product or group differences and decision risk
    • All products must have the same elasticity
  8. What does a sophisticated model alone guarantee about a price estimate?
    • Nothing about causal validity without suitable design and assumptions
    • That the intervention caused the observed sales change
    • That an experiment will be risk-free
    • That the estimate will transport to every product and period