Static Pricing for Single Sample Multi-unit Prophet Inequalities

Pranav Nuti,Peter Westbrook

2024-09-12

Abstract:In this paper, we study $k$-unit single sample prophet inequalities. A seller has $k$ identical, indivisible items to sell. A sequence of buyers arrive one-by-one, with each buyer's private value for the item, $X_i$, revealed to the seller when they arrive. While the seller is unaware of the distribution from which $X_i$ is drawn, they have access to a single sample, $Y_i$ drawn from the same distribution as $X_i$. What strategies can the seller adopt so as to maximize social welfare? Previous work has demonstrated that when $k = 1$, if the seller sets a price equal to the maximum of the samples, they can achieve a competitive ratio of $\frac{1}{2}$ of the social welfare, and recently Pashkovich and Sayutina established an analogous result for $k = 2$. In this paper, we prove that for $k \geq 3$, setting a (static) price equal to the $k^{\text{th}}$ largest sample also obtains a competitive ratio of $\frac{1}{2}$, resolving a conjecture Pashkovich and Sayutina pose. We then consider the situation where $k$ is large. We demonstrate that setting a price equal to the $(k-\sqrt{2k\log k})^{\text{th}}$ largest sample obtains a competitive ratio of $1 - \sqrt{\frac{2\log k}{k}} - o\left(\sqrt{\frac{\log k}{k}}\right)$, and that this is the optimal possible ratio achievable with a static pricing scheme that sets one of the samples as a price.

Computer Science and Game Theory,Data Structures and Algorithms

What problem does this paper attempt to address?

### Problems the paper attempts to solve

This paper mainly studies how the seller can maximize social welfare through a static pricing strategy under the single - sample multi - unit prophet inequalities. The specific problem is described as follows:

1. **Problem background**:
- The seller has \(k\) identical and indivisible items.
- Buyers arrive in sequence, and the private value \(X_i\) of each buyer for the item is revealed to the seller upon arrival.
- The seller does not know the distribution of \(X_i\), but can obtain a single sample \(Y_i\) from the same distribution.

2. **Objective**:
- Design a static pricing strategy so that the seller can maximize social welfare (that is, the sum of the values of all successfully sold items).
- In particular, when \(k\geq3\), can setting the \(k\) - th largest sample as the price achieve a competitive ratio of \(\frac{1}{2}\)?

3. **Previous work**:
- When \(k = 1\), if the seller sets the price as the maximum sample value, a competitive ratio of \(\frac{1}{2}\) can be achieved.
- Recent research shows that when \(k=2\), setting the price as the second - largest sample value can also achieve a competitive ratio of \(\frac{1}{2}\).

4. **Contributions of this paper**:
- It is proved that for \(k\geq3\), setting the price as the \(k\) - th largest sample value can also achieve a competitive ratio of \(\frac{1}{2}\), which solves the conjecture proposed by Pashkovich and Sayutina.
- For larger \(k\), setting the price as the \((k-\sqrt{2k\log k})\) - th largest sample value can achieve a competitive ratio of \(1-\sqrt{\frac{2\log k}{k}}-O\left(\sqrt{\frac{\log k}{k}}\right)\), and this is the best ratio that the static pricing strategy can achieve.

### Formula summary

- **Competitive ratio formula**:
- For \(k\geq3\), the competitive ratio when setting the price as the \(k\) - th largest sample value:
\[
E[\text{ALG}] \geq \frac{1}{2}E\left[\sum_{i = 1}^k X_i\right]
\]
- For larger \(k\), the competitive ratio when setting the price as the \((k-\sqrt{2k\log k})\) - th largest sample value:
\[
E[\text{ALG}] \geq \left(1-\sqrt{\frac{2\log k}{k}}-O\left(\sqrt{\frac{\log k}{k}}\right)\right)E\left[\sum_{i = 1}^k X_i\right]
\]

### Conclusion

Through strict mathematical proofs, this paper shows that under the single - sample multi - unit prophet inequalities, the seller can effectively maximize social welfare through a simple static pricing strategy. In particular, in the case of \(k\geq3\), the previously proposed conjecture is verified, and theoretical support is provided for practical applications.