An official website of the United States government
Large-scale online recommendation systems must facilitate the allocation of a limited number of items among competing users while learning their preferences from user feedback. As a principled way of incorporating market constraints and user incentives in the design, we consider our objectives to be two-fold: maximal social welfare with minimal instability. To maximize social welfare, our proposed framework enhances the quality of recommendations by exploring allocations that optimistically maximize the rewards. To minimize instability, a measure of users' incentives to deviate from recommended allocations, the algorithm prices the items based on a scheme derived from the Walrasian equilibria. Though it is known that these equilibria yield stable prices for markets with known user preferences, our approach accounts for the inherent uncertainty in the preferences and further ensures that the users accept their recommendations under offered prices. To the best of our knowledge, our approach is the first to integrate techniques from combinatorial bandits, optimal resource allocation, and collaborative filtering to obtain an algorithm that achieves sub-linear social welfare regret as well as sub-linear instability. Empirical studies on synthetic and real-world data also demonstrate the efficacy of our strategy compared to approaches that do not fully incorporate all these aspects.
more »
« less
Learning while setting precedents
Abstract A decision maker makes a ruling on a random case in each period. She is uncertain about the correct ruling until conducting a costly investigation. A ruling establishes a precedent, which cannot be violated under binding precedent. We compare the information acquisition incentives, the evolution of standards and the social welfare under nonbinding and binding precedents. Compared to nonbinding precedent, under binding precedent, information acquisition incentives are stronger in earlier periods, but become weaker as more precedents are established. Although erroneous rulings may be perpetuated under binding precedent, welfare can be higher because of the more intensive investigation early on.
more »
« less
- Award ID(s):
- 1730636
- PAR ID:
- 10245905
- Publisher / Repository:
- Wiley-Blackwell
- Date Published:
- Journal Name:
- The RAND Journal of Economics
- Volume:
- 51
- Issue:
- 4
- ISSN:
- 0741-6261
- Page Range / eLocation ID:
- p. 1222-1252
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Moral foundations theory proposes that there are two types of moral domains: the individualizing domain, which relates to individual welfare (comprising harm and fairness foundations), and the binding domain, which relates to communal and spiritual welfare (comprising loyalty, authority, and purity foundations). In this investigation, we demonstrate that this distinction is not universal. Specifically, across five studies (totalN= 1,211) conducted among Jews in Israel and Christians in the United States, we show that the core religious belief that people are created in the image of God is associated not only with purity/divinity values that are typically considered to be part of the binding domain but also with the individualizing moral domain. In two correlational studies, we find that this belief is highly correlated with religiosity but that it predicts greater endorsement of the individualizing moral domain (Studies 1–2). Two experimental studies further establish that this belief is associated with endorsing the individualizing moral domain and the moral foundation of purity, but not the communal foundations (Studies 3–4). Finally, in Study 5, we demonstrate that these experimental findings are not driven by belief in God. We conclude that the distinction between individualizing and binding moral domains is more culturally contingent than previously believed. We discuss the broader implications of the belief in creation in the image of God for understanding moral judgments pertaining to human dignity.more » « less
-
Abstract Motivation is often thought to enhance adaptive decision-making by biasing actions toward rewards and away from punishment. Emerging evidence, however, points to a more nuanced view whereby motivation can both enhance and impair different aspects of decision-making. Model-based approaches have gained prominence over the past decade for developing more precise mechanistic explanations for how incentives impact goal-directed behavior. In this Special Focus, we highlight three studies that demonstrate how computational frameworks help decompose decision processes into constituent cognitive components, as well as formalize when and how motivational factors (e.g., monetary rewards) influence specific cognitive processes, decision-making strategies, and self-report measures. Finally, I conclude with a provocative suggestion based on recent advances in the field: that organisms do not merely seek to maximize the expected value of extrinsic incentives. Instead, they may be optimizing decision-making to achieve a desired internal state (e.g., homeostasis, effort, affect). Future investigation into such internal processes will be a fruitful endeavor for unlocking the cognitive, computational, and neural mechanisms of motivated decision-making.more » « less
-
NA (Ed.)We study the investment incentives created by truthful mechanisms that allocate resources using approximation algorithms. Some approximation algorithms guarantee nearly 100% of the optimal welfare in the allocation problem but guarantee nothing when accounting for investment incentives. An algorithm's allocative and investment guarantees coincide if and only if itsconfirming negative externalitiesare sufficiently small. We introduce fast approximation algorithms for the knapsack problem that have no confirming negative externalities and guarantees close to 100% for both allocation and investment.more » « less
-
Abstract BackgroundEngineers are professionally obligated to protect the safety and well‐being of the public impacted by the technologies they design and maintain. In an increasingly complex sociotechnical world, engineering educators and professional institutions have a duty to train engineers in these responsibilities. Purpose/HypothesisThis article asks, where are engineers trained in their public welfare responsibilities, and how effective is this training? We argue that engineers trained in public welfare responsibilities, especially within engineering education, will demonstrate greater understanding of their duty to recognize and respond to public welfare concerns. We expect training in formal engineering classes to be more broadly impactful than training in contexts like work or professional societies. Data/MethodsWe analyze unique survey data from a representative sample of US practicing engineers using descriptive and regression techniques. ResultsConsistent with expectations, engineers who received public welfare responsibility training in engineering classes are more likely than other engineers to understand their responsibilities to protect public health and safety and problem‐solve collectively, to recognize the importance of social consequences and ethical responsibilities in their own jobs, to have noticed ethical issues in their workplace, and to have taken action about an issue that concerned them. Training through other parts of college, workplaces, or professional societies has comparatively little impact. Concerningly, nearly a third of engineers reported never being trained in public welfare responsibilities. ConclusionThese results suggest that training in engineering education can shape engineers' long‐term understanding of their public welfare responsibilities. They underscore the need for these responsibilities to be taught as a core, non‐negotiable part of engineering education.more » « less
