Pandora's Box Problem Over Time

Georgios Amanatidis,Federico Fusco,Rebecca Reiffenhäuser,Artem Tsikiridis

2024-07-22

Abstract:The Pandora's Box problem models the search for the best alternative when evaluation is costly. In its simplest variant, a decision maker is presented with $n$ boxes, each associated with a cost of inspection and a distribution over the reward hidden within. The decision maker inspects a subset of these boxes one after the other, in a possibly adaptive ordering, and obtains as utility the difference between the largest reward uncovered and the sum of the inspection costs. While this classic version of the problem is well understood (Weitzman 1979), recent years have seen a flourishing of the literature on variants of the problem. In this paper, we introduce a general framework -- the Pandora's Box Over Time problem -- that captures a wide range of variants where time plays a role, e.g., as it might constrain the schedules of exploration and influence both costs and rewards. In the Pandora's Box Over Time problem, each box is characterized by time-dependent rewards and costs, and inspecting it might require a box-specific processing time. Moreover, once a box is inspected, its reward may deteriorate over time, possibly differently for each box. Our main result is an efficient $21.3$-approximation to the optimal strategy, which is NP-hard to compute in general. We further obtain improved results for the natural special cases where boxes have no processing time, or when costs and reward distributions do not depend on time (but rewards may deteriorate after inspecting).

Computer Science and Game Theory,Data Structures and Algorithms,Probability

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to find the optimal balance between cost and benefit after introducing the time factor on the basis of the classic Pandora's Box problem. Specifically, decision - makers face multiple "boxes", and each box has an inspection cost and a potential reward. Different from the classic version, in this problem, **time and inspection order have an impact on cost and reward**. For example, some boxes may only be inspected within a specific time period, or the reward after inspection may decrease over time.

### Core challenges of the problem
1. **Time - dependent cost and reward**: The inspection cost and reward distribution of each box are no longer fixed but change over time.
2. **Processing time**: After inspecting one box, it may be necessary to wait for a while before inspecting the next box.
3. **Reward depreciation**: Once a box is inspected, its reward may decrease over time, and different boxes may have different depreciation rates.

### Research objectives
The authors propose a new framework - **Pandora’s Box Over Time** to capture these time - related changes and study the following questions under this framework:
- How to design a strategy to maximize the net benefit of the maximum final reward minus the total inspection cost when considering time and cost?
- Can effective approximate solutions be provided for such complex time - dependent problems?

### Main contributions
1. **Propose a general framework** that can cover multiple time - dependent variants.
2. **Provide an efficient approximate algorithm** that can calculate a 21.3 - times approximately optimal strategy within polynomial time.
3. **Improve for special cases**, such as when the processing time of all boxes is zero, or when the cost and reward distribution do not depend on time (but the reward may be depreciated), better approximate results are provided.

Through these contributions, the authors not only extend the classic Pandora's Box problem but also provide more practical solutions for practical application scenarios (such as recruitment, house purchase, etc.).