Events & Seminars

We study the assortment optimization problem and show that local optima are global optima for all discrete choice models that can be represented by the Markov Chain model. We develop a forward greedy heuristic that finds an optimal assortment for the Markov Chain model and runs in $O(n^2)$ iterations. The heuristic has performance bound $1/n$ for any regular choice model which is best possible among polynomial heuristics. We also propose a backward greedy heuristic that is optimal for Markov chain model and requires fewer iterations. Numerical results show that our heuristics performs significantly better than the estimate then optimize method and the revenue-ordered assortment heuristic when the ground truth is a latent class multinomial logit choice model. Based on the greedy heuristics, we develop a learning algorithm that enjoys asymptotic optimal regret for the Markov chain choice model and avoids parameter estimations, focusing instead on binary comparisons of revenues.

Many operations problems are associated with some form of a prediction problem. For instance, one cannot solve a supply chain problem without predicting demand. One cannot solve a shortest path problem without predicting travel times. One cannot solve a personalized pricing problem without predicting consumer valuations. In each of these problems, each instance is characterized by a context (or features). For instance, demand depends on prices and trends, travel times depend on weather and holidays, and consumer valuations depend on user demographics and click history. In this talk, we review recent results on how to solve such contextual optimization problems, with a particular emphasis on techniques that blend the prediction and decision tasks together.

Emission permits are widely adopted to combat climate change and regulatory authorities sometimes allow for the inter-temporal banking and borrowing of emission permits so that firms can flexibly respond to market uncertainties. We find that such time flexibility may lead to poor social performance, especially when the production cost fluctuation is sufficiently large. This result is failed to be captured by the classic simplified assumption where firms cannot sub-exercise emission permits. Furthermore, we demonstrate that the inter-temporal permits transfer should be prohibited when the market is at the red ocean stage or when the pollutant generated relatively instant damage. Lastly, we analyze some restricted permits transfer policies such as transfer discount and transfer cap, which are shown to dominate both the taxation and the non-transferable permits in terms of social welfare.

Information technology (IT) professionals play an important role in firms' IT investments, innovation, and entrepreneurship, contributing to significant economic growth in the U.S. The use of temporary work visas and related immigration policies has attracted a significant controversy and policy debates in the U.S. On the one hand, foreign IT professionals complement domestic IT professionals by facilitating innovation and entrepreneurship. On the other hand, the foreign IT professionals substitute the domestic counterparts by intensifying labor market competition. In this study, we focus on an extension in the Optional Practical Training (OPT) program for STEM graduates from U.S. institutions. Specifically, we explore the effects of the OPT extension on the number and wage of domestic workers in STEM occupations and how these effects differ between IT and non-IT STEM occupations. Our results demonstrate that an increase in the supply of foreign IT professionals from the OPT extension boosts the employment of domestic IT professionals. This study contributes to the information systems, labor economics, and public policy literature by quantifying the impacts of a policy change on the employment of IT professionals and provides rich implications for policymakers.

Even when human point forecasts are less accurate than data-based algorithm predictions, they can still help boost performance by being used as algorithm inputs. Assuming one uses human judgment indirectly in this manner, we propose changing the elicitation question from the traditional direct forecast (DF) to what we call the private information adjustment (PIA): how much the human thinks the algorithm should adjust its forecast to account for information the human has that is unused by the algorithm. Using stylized models with and without random error, we theoretically prove that human random error makes eliciting the PIA lead to more accurate predictions than eliciting the DF. However, this DF-PIA gap does not exist for perfectly consistent forecasters. The DF-PIA gap is increasing in the random error that people make while incorporating public information (data that the algorithm uses) but is decreasing in the random error that people make while incorporating private information (data that only the human can use). In controlled experiments with students and Amazon Mechanical Turk workers, we find support for these hypotheses.

This paper highlights two channels through which blockchain-enabled tokenization can alleviate moral hazard frictions between founders, investors, and users of a platform: token financing and decentralized governance. We consider an entrepreneur who uses outside financing and exerts private effort to build a platform, and users who decide whether to join in response to the platform’s dynamic transaction fee policy. We first show that raising capital by issuing tokens rather than equity mitigates effort under-provision because the payoff to equity investors depends on profit, whereas the payoff to token investors depends on transaction volume, which is less sensitive to effort. Second, we show that decentralized governance associated with tokenization eliminates a potential holdup of platform users, which in turn alleviates the need to provide users with incentives to join, reducing the entrepreneur’s financing burden. The downside of tokenization is that it puts a cap on how much capital the entrepreneur can raise. Namely, if tokens are highly liquid, i.e., they change hands many times per unit of time, their market capitalization is small relative to the NPV of the platform profits, limiting how much money one can raise by issuing tokens rather than equity. If building the platform is expensive, this can distort the capacity investment. The resulting trade-off between the benefits and costs of tokenization leads to several predictions regarding adoption.

Prophet inequalities are a useful tool for designing online allocation procedures and comparing their performance to the optimal offline allocation. In the basic setting of $k$-unit prophet inequalities, the magical procedure of Alaei (2011) with its celebrated performance guarantee of $1-1/sqrt(k+3)$ has found widespread adoption in mechanism design and general online allocation problems in online advertising, healthcare scheduling, and revenue management. Despite being commonly used for implementing a fractional allocation in an online fashion, the tightness of Alaei’s procedure for a given $k$ has remained unknown. In this paper we resolve this question, characterizing the tight bound by identifying the structure of the optimal online implementation, and consequently improving the best-known guarantee for $k$-unit prophet inequalities for all $k>1$.

We also consider the more general online stochastic knapsack problem where each individual allocation can consume an arbitrary fraction of the initial capacity. Here we introduce a new “best-fit” procedure for implementing a fractionally-feasible knapsack solution online, with a performance guarantee of $1/( 3+ e^(-2) ) ~ 0.319$, which we also show is tight with respect to the standard LP relaxation. This improves the previously best- known guarantee of 0.2 for online knapsack.

Our analysis differs from existing ones by eschewing the need to split items into “large” or “small” based on capacity consumption, using instead an invariant for the overall utilization on different sample paths.

Finally, we refine our technique for the unit-density special case of knapsack, and improve the guarantee from 0.321 to 0.3557 in the multi-resource appointment scheduling application of Stein et al. (2020).

(Joint work with Jiashuo Jiang and Jiawei Zhang)

This paper examines an innovative return policy, return insurance, emerging on various shopping platforms such as Taobao.com and JD.com. Return insurance is underwritten by an insurer and can be purchased by either a retailer or a consumer. Under such insurance, the insurer partially compensates consumers for their hassle costs associated with product return. We analyze the informational roles of return insurance when product quality is the retailer's private information, consumers infer quality from the retailer's price and insurance adoption, and the insurer strategically chooses insurance premiums.

We show that return insurance can be an effective signal of high quality. When consumers have little confidence about high quality and expect a significant gap between high and low qualities, a high-quality retailer can be differentiated from a low-quality retailer solely through its adoption of return insurance. We confirm, both analytically and empirically with a data set consisting of over 10,000 sellers on JD.com, that return insurance is more likely adopted by higher-quality sellers under information asymmetry. Furthermore, we find that the presence of the third party (i.e., the insurer) leads to double marginalization in signaling, which strengthens a signal's differentiating power and sometimes renders return insurance a preferred signal, in comparison with free return, whereby retailers directly compensate for consumers' return hassles. As an effective and costly signal of quality, return insurance may also improve consumer surplus and reduce product returns. Its profit advantage to the insurer is most pronounced under significant quality uncertainty.

Traffic congestion is a serious global issue. A potential solution, which requires zero investment in infrastructure, is to convince solo car users to carpool. In this paper, we leverage the Waze Carpool service and run the largest ever digital field experiment to nudge commuters to carpool. Our field experiment involves more than half a million users across four U.S. states between June 10 and July 3, 2019. We identify users who can save a significant commute time by carpooling through the use of a high- occupancy vehicle (HOV) lane, users who can still use a HOV lane but have a low time saving, and users who do not have access to a HOV lane on their commute. We send them in-app notifications with different framings: mentioning the HOV lane, highlighting the time saving, emphasizing the monetary welcome bonus (for users who do not have access to a HOV lane), and a generic carpool invitation. We find a strong relationship between the affinity to carpool and the potential time saving through a HOV lane. Specifically, we estimate that mentioning the HOV lane increases the click-through rate (i.e., proportion of users who clicked on the button inviting them to try the carpool service) and the on-boarding rate (i.e., proportion of users who signed up and created an account with the carpool service) by 133-185% and 64-141%, respectively relative to a generic invitation. We conclude by discussing the implications of our findings for carpool platforms and public policy.

(Joint work with Michael-David Fiszer, Avia Ratzon, and Roy Sasson)