ERN: Statistical Decision Theory; Operations Research (Topic)最新文献

{"title":"Market Efficiency in the Age of Big Data","authors":"Ian Martin, S. Nagel","doi":"10.3386/w26586","DOIUrl":"https://doi.org/10.3386/w26586","url":null,"abstract":"Modern investors face a high-dimensional prediction problem: thousands of observable variables are potentially relevant for forecasting. We reassess the conventional wisdom on market efficiency in light of this fact. In our model economy, which resembles a typical machine learning setting, N assets have cash flows that are a linear function of J firm characteristics, but with uncertain coefficients. Risk-neutral Bayesian investors impose shrinkage (ridge regression) or sparsity (Lasso) when they estimate the J coefficients of the model and use them to price assets. When J is comparable in size to N, returns appear cross-sectionally predictable using firm characteristics to an econometrician who analyzes data from the economy ex post. A factor zoo emerges even without p-hacking and data-mining. Standard in-sample tests of market efficiency reject the no-predictability null with high probability, despite the fact that investors optimally use the information available to them in real time. In contrast, out-of-sample tests retain their economic meaning.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129636671","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Existence and Uniqueness of Solutions for a Partially Observed Stochastic Control Problem","authors":"A. Bensoussan, M. Çakanyıldırım, M. Li, S. Sethi","doi":"10.2139/ssrn.3493943","DOIUrl":"https://doi.org/10.2139/ssrn.3493943","url":null,"abstract":"We develop a general methodology for a partially observed stochastic control problem. The dynamics is governed by a discrete-time Markov process. We describe an application to an inventory system with possibility of shrinkage, and introduce unnormalized conditional probabilities to transform the nonlinear state evolution into a linear one. We then prove the existence and uniqueness of the solution for the Bellman equation between the floor and the ceiling functions in the case of unbounded costs.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133166277","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diffusion Approximations for a Class of Sequential Experimentation Problems","authors":"V. F. Araman, René Caldentey","doi":"10.2139/ssrn.3479676","DOIUrl":"https://doi.org/10.2139/ssrn.3479676","url":null,"abstract":"A decision maker (DM) must choose an action in order to maximize a reward function that depends on the DM’s action as well as on an unknown parameter Θ. The DM can delay taking the action in order to experiment and gather additional information on Θ. We model the problem using a Bayesian sequential experimentation framework and use dynamic programming and diffusion-asymptotic analysis to solve it. For that, we consider environments in which the average number of experiments that is conducted per unit of time is large and the informativeness of each individual experiment is low. Under such regimes, we derive a diffusion approximation for the sequential experimentation problem, which provides a number of important insights about the nature of the problem and its solution. First, it reveals that the problems of (i) selecting the optimal sequence of experiments to use and (ii) deciding the optimal time when to stop experimenting decouple and can be solved independently. Second, it shows that an optimal experimentation policy is one that chooses the experiment that maximizes the instantaneous volatility of the belief process. Third, the diffusion approximation provides a more mathematically malleable formulation that we can solve in closed form and suggests efficient heuristics for the nonasympototic regime. Our solution method also shows that the complexity of the problem grows only quadratically with the cardinality of the set of actions from which the decision maker can choose. We illustrate our methodology and results using a concrete application in the context of assortment selection and new product introduction. Specifically, we study the problem of a seller who wants to select an optimal assortment of products to launch into the marketplace and is uncertain about consumers’ preferences. Motivated by emerging practices in e-commerce, we assume that the seller is able to use a crowd voting system to learn these preferences before a final assortment decision is made. In this context, we undertake an extensive numerical analysis to assess the value of learning and demonstrate the effectiveness and robustness of the heuristics derived from the diffusion approximation. This paper was accepted by Omar Besbes, revenue management and market analytics.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131070738","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Estimating Certain Integral Probability Metrics (IPMs) Is as Hard as Estimating under the IPMs","authors":"Tengyuan Liang","doi":"10.2139/ssrn.3714012","DOIUrl":"https://doi.org/10.2139/ssrn.3714012","url":null,"abstract":"We study the minimax optimal rates for estimating a range of Integral Probability Metrics (IPMs) between two unknown probability measures, based on $n$ independent samples from them. Curiously, we show that estimating the IPM itself between probability measures, is not significantly easier than estimating the probability measures under the IPM. We prove that the minimax optimal rates for these two problems are multiplicatively equivalent, up to a $log log (n)/log (n)$ factor.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131819344","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimal Subscription Planning for Digital Goods","authors":"S. Alaei, A. Makhdoumi, Azarakhsh Malekian","doi":"10.2139/ssrn.3476296","DOIUrl":"https://doi.org/10.2139/ssrn.3476296","url":null,"abstract":"Alaei et al. consider the problem of a streaming platform that offers different set of contents with different subscription fees with the goal of maximizing revenue. They characterize the necessary and sufficient condition for the consumer preferences under which a single-tier subscription is optimal. In particular, they show the preferences of the consumers for different contents must be aligned for a single-tier subscription to be optimal. This condition is often violated in practice, establishing the suboptimality of a single-tier offering. They further show that a subscription fee proportional to the number of contents is near-optimal.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127849326","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Optimal Spatial Pricing for an On-demand Ride-sourcing Service Platform","authors":"X. Chen, Chen Chuqiao, Weijun Xie","doi":"10.2139/ssrn.3464228","DOIUrl":"https://doi.org/10.2139/ssrn.3464228","url":null,"abstract":"This paper investigates a long-term optimal spatial pricing strategy for a ride-sourcing platform that serves a particular (possibly populated) area with profit-driven service providers (i.e., drivers) and time- and price- sensitive customers. By observing that oftentimes, the pricing strategy is anisotropic and spatially dependent, and both the supply and request rates are endogenous, we build an analytical bi-level optimization model. In the upper-level formulation, the ride-sourcing platform aims to set up the spatially heterogeneous pricing strategy to maximize its total profit. While in the lower level, we solve the trip distribution model that characterizes the optimal flow rates among zones given the travel demand rate at each zone. We prove that when the platform seeks to expand its business, the optimal number of participating drivers and their optimal wages will be influenced not only by the pricing strategy but also by the levels of service of the service zones. Our further investigation shows that the profit at a particular zone can be also influenced by the potential customers' service requests from the other zones. Finally, we use the real-world data provided by DiDi Chuxing to numerically validate our theoretical results.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"96 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117314606","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"On Demand Uncertainty in the Newsvendor Model","authors":"R. Butters","doi":"10.2139/ssrn.3461096","DOIUrl":"https://doi.org/10.2139/ssrn.3461096","url":null,"abstract":"I provide three comparative statics involving the level of demand uncertainty for the newsvendor model, two of which lead to robust predictions. I show that for distributions of demand that are greater in the dispersive order, both the expected (censored) sales and share of inventory sold fall. These monotone comparative statics occur despite the lack of one for the (optimal) inventory choice.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127062786","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Dynamic Revenue and Project Management","authors":"Hossein Jahandideh","doi":"10.2139/ssrn.3453841","DOIUrl":"https://doi.org/10.2139/ssrn.3453841","url":null,"abstract":"We consider a service provider who makes dynamic pricing decisions for deadline-based projects with specific resource requirements. The provider receives service requests and quotes a menu of prices for different completion times, based on the project type and resources required to complete the project. Given the fixed resource capacity per period, the provider has the flexibility of processing a project in different time periods as long as all projects are completed by their respective deadlines. This flexibility adds a dimension of dynamic project scheduling to the provider's decision, making it a non-traditional dynamic network revenue management (DNRM) problem. We first present a procedure to reformulate this problem as a traditional DNRM and show that the reformulation is equivalent to the original problem. We then consider the dynamic problem of setting prices for different completion times and present a fast, effective, and scalable procedure to approximately solve the problem. The approximate solution provides a real-time pricing strategy as well as an upper bound on the optimal revenue for the purpose of evaluation. We discuss how to extend the procedure to general choice models under fairly mild assumptions. We numerically test the procedure under various choice models and parameter settings and demonstrate its effectiveness.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121196676","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Crowdsourcing Last-Mile Deliveries","authors":"Soraya Fatehi, M. Wagner","doi":"10.2139/ssrn.3434625","DOIUrl":"https://doi.org/10.2139/ssrn.3434625","url":null,"abstract":"Problem definition: Because of the emergence and development of e-commerce, customers demand faster and cheaper delivery services. However, many retailers find it challenging to efficiently provide fast and on-time delivery services to their customers. Academic/practical relevance: Amazon and Walmart are among the retailers that are relying on independent crowd drivers to cope with on-demand delivery expectations. Methodology: We propose a novel robust crowdsourcing optimization model to study labor planning and pricing for crowdsourced last-mile delivery systems that are utilized for satisfying on-demand orders with guaranteed delivery time windows. We develop our model by combining crowdsourcing, robust queueing, and robust routing theories. We show the value of the robust optimization approach by analytically studying how to provide fast and guaranteed delivery services utilizing independent crowd drivers under uncertainties in customer demands, crowd availability, service times, and traffic patterns; we also allow for trend and seasonality in these uncertainties. Results: For a given delivery time window and an on-time delivery guarantee level, our model allows us to analytically derive the optimal delivery assignments to available independent crowd drivers and their optimal hourly wage. Our results show that crowdsourcing can help firms decrease their delivery costs significantly while keeping the promise of on-time delivery to their customers. Managerial implications: We provide extensive managerial insights and guidelines for how such a system should be implemented in practice.","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123243896","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Product Design Enhancement for Fashion Retailing","authors":"Yiwei Wang, Vidyanand Choudhary, Shuya Yin","doi":"10.2139/ssrn.3392176","DOIUrl":"https://doi.org/10.2139/ssrn.3392176","url":null,"abstract":"As the fashion industry increasingly embraces artificial intelligence (AI), we investigate how a fast-fashion retailer should choose between using a manual design strategy or an AI-assisted design strategy to enhance existing products. A manual design is a traditional and basic approach that involves human designers only, whereas an AI-assisted design is a more innovative approach that involves both human designers and AI technologies. In this paper, the overall product enhancement is measured by two key attributes: product quality and product trendiness. Product quality can be measured by the product’s longevity as reflected by the quality of the materials and types of fabric and stitching used, where the product’s improvement level can be determined by the retailer in a continuous range. Consequently, the retailer may choose different levels of product quality under different design strategies. The two design approaches also lead to different natures of product trendiness, which is reflected by features such as styles, new materials, and colors, to name just a few. Specifically, we assume that the traditional manual design can predict well how trendy or popular the new product is. Hence, the trendiness attribute under the manual design is deterministic. However, given the uncertain nature of the AI-assisted design technology and the needed coordination between human designers and the adopted technologies, the trendiness of the new product designed under the AI-assisted approach is assumed uncertain. Two sets of designing costs are considered in product enhancement: the fixed design cost that is irrespective of the production volume and the variable marginal cost. Our analysis of the base model highlights the importance of decomposing different costs in determining the optimal design strategy. Specifically, the manual design is preferred when the fixed cost carries more weight, whereas the AI-assisted design is preferred when the marginal cost is a more important factor. Moreover, a higher level of innovation uncertainty under the AI-assisted design gives this strategy an advantage over the manual design. In our extended models, we demonstrate that (1) these results are robust even if the retailer does not have the flexibility to offer the existing product when the AI-assisted design is unpopular, and (2) the relative position of human designers in the two design approaches has an impact on the effects of these costs. Supplemental Material: The online appendix is available at https://doi.org/10.1287/serv.2023.0315 .","PeriodicalId":200007,"journal":{"name":"ERN: Statistical Decision Theory; Operations Research (Topic)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-07-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125095948","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}