arXiv subjects

Foster Provost

Publications and source records attributed to Foster Provost.

17 recordsLinked to original sources

arXiv2026

AI Strategy: How to Choose What AI Product to Implement

Firms struggle to choose AI projects that pay off: two projects can look equally promising to smart, motivated stakeholders and yet deserve opposite decisions. At the residential real-estate brokerage Compass, one AI product (Likely-to-Sell recommendations) flagged sales outreach opportunities and went on to account for nine figures in annual gross commission revenue. Another championed AI product (a Time-on-Market pricing tool) was rightly shelved. A simple ROI estimate could not distinguish the two. We present expected ROI (eROI), a framework that decomposes each bet into three components and rates them separately: Value if Successful, Likelihood of Success, and Investment Required. Each maps to a question executives can answer before building: How valuable would it be if it worked? How likely is it to work? And what would it cost to implement? Separating the three breaks a common catch-22: teams cannot estimate ROI until they know whether a project will work, yet cannot know whether it will work without building it. Judging Value if Successful on its own dissolves the loop, letting a team argue that a product would be valuable if it worked while it weighs how likely that is. The framework also asks, before ranking anything, whether there are enough good ideas on the table. After ranking, it guides assembling a portfolio of bets rather than funding only the single top-ranked project. We illustrate eROI on Compass's candidate AI products. Precise ROI estimates are hard to make given the inherent uncertainty of AI projects. Coarse business-level ratings of the three components are enough to tell strong bets from weak ones.

cs.CY↗

arXiv2026

Logging Policy Design for Off-Policy Evaluation

Off-policy evaluation (OPE) estimates the value of a target treatment policy (e.g., a recommender system) using data collected by a different logging policy. It enables high-stakes experimentation without live deployment, yet in practice accuracy depends heavily on the logging policy used to collect data for computing the estimate. We study how to design logging policies that minimize OPE error for given target policies. We characterize a fundamental reward-coverage tradeoff: concentrating probability mass on high-reward actions reduces variance but risks missing signal on actions the target policy may take. We propose a unifying framework for logging policy design and derive optimal policies in canonical informational regimes where the target policy and reward distribution are (i) known, (ii) unknown, and (iii) partially known through priors or noisy estimates at logging time. Our results provide actionable guidance for firms choosing among multiple candidate recommendation systems. We demonstrate the importance of treatment selection when gathering data for OPE, and describe theoretically optimal approaches when this is a firm's primary objective. We also distill practical design principles for selecting logging policies when operational constraints prevent implementing the theoretical optimum.

stat.ML↗

arXiv2026

Prompt-Counterfactual Explanations for Generative AI System Behavior

As generative AI systems become integrated into real-world applications, organizations increasingly need to be able to understand and interpret their behavior. In particular, decision-makers need to understand what causes generative AI systems to exhibit specific output characteristics. Within this general topic, this paper examines a key question: what is it about the input -- the prompt -- that causes an LLM-based generative AI system to produce output that exhibits specific characteristics, such as toxicity, negative sentiment, or political bias. To examine this question, we adapt a common technique from the Explainable AI literature: counterfactual explanations. We explain why traditional counterfactual explanations cannot be applied directly to generative AI systems, due to several differences in how generative AI systems function. We then propose a flexible framework that adapts counterfactual explanations to non-deterministic, generative AI systems in scenarios where downstream classifiers can reveal key characteristics of their outputs. Based on this framework, we introduce an algorithm for generating prompt-counterfactual explanations (PCEs). Finally, we demonstrate the production of counterfactual explanations for generative AI systems with three case studies, examining different output characteristics (viz., political leaning, toxicity, and sentiment). The case studies further show that PCEs can streamline prompt engineering to suppress undesirable output characteristics and can enhance red-teaming efforts to uncover additional prompts that elicit undesirable outputs. Ultimately, this work lays a foundation for prompt-focused interpretability in generative AI: a capability that will become indispensable as these models are entrusted with higher-stakes tasks and subject to emerging regulatory requirements for transparency and accountability.

cs.LG↗

arXiv2025

Beware of "Explanations" of AI

Understanding the decisions made and actions taken by increasingly complex AI system remains a key challenge. This has led to an expanding field of research in explainable artificial intelligence (XAI), highlighting the potential of explanations to enhance trust, support adoption, and meet regulatory standards. However, the question of what constitutes a "good" explanation is dependent on the goals, stakeholders, and context. At a high level, psychological insights such as the concept of mental model alignment can offer guidance, but success in practice is challenging due to social and technical factors. As a result of this ill-defined nature of the problem, explanations can be of poor quality (e.g. unfaithful, irrelevant, or incoherent), potentially leading to substantial risks. Instead of fostering trust and safety, poorly designed explanations can actually cause harm, including wrong decisions, privacy violations, manipulation, and even reduced AI adoption. Therefore, we caution stakeholders to beware of explanations of AI: while they can be vital, they are not automatically a remedy for transparency or responsible AI adoption, and their misuse or limitations can exacerbate harm. Attention to these caveats can help guide future research to improve the quality and impact of AI explanations.

cs.LG↗

arXiv2024

The Illusion of Collusion

Algorithmic agents are used in a variety of competitive decision-making settings, including pricing contexts that range from online retail to residential home rental. We study the emergence of algorithmic collusion when competing agents employ multi-armed bandit algorithms and competition is modeled as a repeated Prisoner's Dilemma game. Notably, agents in our setting perform online learning with no prior model of game structure and have no direct knowledge of competitor states or actions, thus they cannot learn strategies that depend on these factors. These context-free bandits nonetheless frequently learn seemingly collusive behavior, a phenomenon we term naive collusion. Our results reveal that whether naive collusion emerges depends starkly on the choice of behavior policy employed by bandit learners. The mechanism underpinning the emergence of collusive outcomes is synchronicity in agent action plays, where synchronicity captures how often agents play the same action. We show that in the long-run, naive algorithmic collusion never emerges when both agents use a broad class of persistently random algorithms, including the epsilon-greedy algorithm without epsilon decay, sometimes emerges when both agents use greedy-in-the-limit algorithms which feature randomness during exploration but are asymptotically deterministic, and always emerges when both agents use deterministic bandit learning algorithms like those in the well-known upper confidence bound (UCB) family. We highlight market and algorithmic conditions under which one can and cannot predict a priori whether collusion will occur. Our findings have several policy implications: preventing pricing algorithms from conditioning their actions on competitor prices may not preclude algorithmic collusion, symmetry in algorithms may increase collusion potential, and the emergence of algorithmic collusion is path dependent.

econ.GN↗

arXiv2024

Causal Post-Processing of Predictive Models

Organizations increasingly rely on predictive models to decide who should be targeted for interventions, such as marketing campaigns, customer retention offers, or medical treatments. Yet these models are usually built to predict outcomes (e.g., likelihood of purchase or churn), not the actual impact of an intervention. As a result, the scores (predicted values) they produce are often imperfect guides for allocating resources. Causal effects can be estimated with randomized experiments, but experiments are costly, limited in scale, and tied to specific actions. We propose causal post-processing (CPP), a family of techniques that uses limited experimental data to refine the outputs of predictive models, so they better align with causal decision making. The CPP family spans approaches that trade off flexibility against data efficiency, unifying existing methods and motivating new ones. Through simulations and an empirical study in digital advertising, we show that CPP can improve intervention decisions, particularly when predictive models capture a useful but imperfect causal signal. Our results show how organizations can combine predictive modeling with experimental evidence to make more effective and scalable intervention decisions.

stat.ML↗

arXiv2023

The Impact of Cloaking Digital Footprints on User Privacy and Personalization

Our online lives generate a wealth of behavioral records -'digital footprints'- which are stored and leveraged by technology platforms. This data can be used to create value for users by personalizing services. At the same time, however, it also poses a threat to people's privacy by offering a highly intimate window into their private traits (e.g., their personality, political ideology, sexual orientation). Prior work has proposed a potential remedy: The cloaking of users' footprints. That is, platforms could allow users to hide portions of their digital footprints from predictive algorithms to avoid undesired inferences. While such an approach has been shown to offer privacy protection in the moment, there are two open questions. First, it remains unclear how well cloaking performs over time. As people constantly leave new digital footprints, the algorithm might regain the ability to predict previously cloaked traits. Second, cloaking digital footprints to avoid one undesirable inference may degrade the performance of models for other, desirable inferences (e.g., those driving desired personalized content). In the light of these research gaps, our contributions are twofold: 1) We propose a novel cloaking strategy that conceals 'metafeatures' (automatically generated higher-level categories) and compares its effectiveness against existing cloaking approaches, and 2) we test the spill-over effects of cloaking one trait on the accuracy of inferences on other traits. A key finding is that the effectiveness of cloaking degrades over times, but the rate at which it degrades is significantly smaller when cloaking metafeatures rather than individual footprints. In addition, our findings reveal the expected trade-off between privacy and personalization: Cloaking an undesired trait also partially conceals other desirable traits.

cs.CY↗

arXiv2021

Causal Decision Making and Causal Effect Estimation Are Not the Same... and Why It Matters

Causal decision making (CDM) based on machine learning has become a routine part of business. Businesses algorithmically target offers, incentives, and recommendations to affect consumer behavior. Recently, we have seen an acceleration of research related to CDM and causal effect estimation (CEE) using machine-learned models. This article highlights an important perspective: CDM is not the same as CEE, and counterintuitively, accurate CEE is not necessary for accurate CDM. Our experience is that this is not well understood by practitioners or most researchers. Technically, the estimand of interest is different, and this has important implications both for modeling and for the use of statistical models for CDM. We draw on prior research to highlight three implications. (1) We should consider carefully the objective function of the causal machine learning, and if possible, optimize for accurate treatment assignment rather than for accurate effect-size estimation. (2) Confounding does not have the same effect on CDM as it does on CEE. The upshot is that for supporting CDM it may be just as good or even better to learn with confounded data as with unconfounded data. Finally, (3) causal statistical modeling may not be necessary to support CDM because a proxy target for statistical modeling might do as well or better. This third observation helps to explain at least one broad common CDM practice that seems wrong at first blush: the widespread use of non-causal models for targeting interventions. The last two implications are particularly important in practice, as acquiring (unconfounded) data on all counterfactuals can be costly and often impracticable. These observations open substantial research ground. We hope to facilitate research in this area by pointing to related articles from multiple contributing fields, including two dozen articles published the last three to four years.

stat.ML↗

arXiv2020

A Comparison of Methods for Treatment Assignment with an Application to Playlist Generation

This study presents a systematic comparison of methods for individual treatment assignment, a general problem that arises in many applications and has received significant attention from economists, computer scientists, and social scientists. We group the various methods proposed in the literature into three general classes of algorithms (or metalearners): learning models to predict outcomes (the O-learner), learning models to predict causal effects (the E-learner), and learning models to predict optimal treatment assignments (the A-learner). We compare the metalearners in terms of (1) their level of generality and (2) the objective function they use to learn models from data; we then discuss the implications that these characteristics have for modeling and decision making. Notably, we demonstrate analytically and empirically that optimizing for the prediction of outcomes or causal effects is not the same as optimizing for treatment assignments, suggesting that in general the A-learner should lead to better treatment assignments than the other metalearners. We demonstrate the practical implications of our findings in the context of choosing, for each user, the best algorithm for playlist generation in order to optimize engagement. This is the first comparison of the three different metalearners on a real-world application at scale (based on more than half a billion individual treatment assignments). In addition to supporting our analytical findings, the results show how large A/B tests can provide substantial value for learning treatment assignment policies, rather than simply choosing the variant that performs best on average.

econ.EM↗

arXiv2020

Explaining Data-Driven Decisions made by AI Systems: The Counterfactual Approach

We examine counterfactual explanations for explaining the decisions made by model-based AI systems. The counterfactual approach we consider defines an explanation as a set of the system's data inputs that causally drives the decision (i.e., changing the inputs in the set changes the decision) and is irreducible (i.e., changing any subset of the inputs does not change the decision). We (1) demonstrate how this framework may be used to provide explanations for decisions made by general, data-driven AI systems that may incorporate features with arbitrary data types and multiple predictive models, and (2) propose a heuristic procedure to find the most useful explanations depending on the context. We then contrast counterfactual explanations with methods that explain model predictions by weighting features according to their importance (e.g., SHAP, LIME) and present two fundamental reasons why we should carefully consider whether importance-weight explanations are well-suited to explain system decisions. Specifically, we show that (i) features that have a large importance weight for a model prediction may not affect the corresponding decision, and (ii) importance weights are insufficient to communicate whether and how features influence decisions. We demonstrate this with several concise examples and three detailed case studies that compare the counterfactual approach with SHAP to illustrate various conditions under which counterfactual explanations explain data-driven decisions better than importance weights.

cs.LG↗

arXiv2019

Counterfactual Explanation Algorithms for Behavioral and Textual Data

We study the interpretability of predictive systems that use high-dimensonal behavioral and textual data. Examples include predicting product interest based on online browsing data and detecting spam emails or objectionable web content. Recently, counterfactual explanations have been proposed for generating insight into model predictions, which focus on what is relevant to a particular instance. Conducting a complete search to compute counterfactuals is very time-consuming because of the huge dimensionality. To our knowledge, for behavioral and text data, only one model-agnostic heuristic algorithm (SEDC) for finding counterfactual explanations has been proposed in the literature. However, there may be better algorithms for finding counterfactuals quickly. This study aligns the recently proposed Linear Interpretable Model-agnostic Explainer (LIME) and Shapley Additive Explanations (SHAP) with the notion of counterfactual explanations, and empirically benchmarks their effectiveness and efficiency against SEDC using a collection of 13 data sets. Results show that LIME-Counterfactual (LIME-C) and SHAP-Counterfactual (SHAP-C) have low and stable computation times, but mostly, they are less efficient than SEDC. However, for certain instances on certain data sets, SEDC's run time is comparably large. With regard to effectiveness, LIME-C and SHAP-C find reasonable, if not always optimal, counterfactual explanations. SHAP-C, however, seems to have difficulties with highly unbalanced data. Because of its good overall performance, LIME-C seems to be a favorable alternative to SEDC, which failed for some nonlinear models to find counterfactuals because of the particular heuristic search algorithm it uses. A main upshot of this paper is that there is a good deal of room for further research. For example, we propose algorithmic adjustments that are direct upshots of the paper's findings.

cs.AI↗

arXiv2017

Big Data, Data Science, and Civil Rights

Advances in data analytics bring with them civil rights implications. Data-driven and algorithmic decision making increasingly determine how businesses target advertisements to consumers, how police departments monitor individuals or groups, how banks decide who gets a loan and who does not, how employers hire, how colleges and universities make admissions and financial aid decisions, and much more. As data-driven decisions increasingly affect every corner of our lives, there is an urgent need to ensure they do not become instruments of discrimination, barriers to equality, threats to social justice, and sources of unfairness. In this paper, we argue for a concrete research agenda aimed at addressing these concerns, comprising five areas of emphasis: (i) Determining if models and modeling procedures exhibit objectionable bias; (ii) Building awareness of fairness into machine learning methods; (iii) Improving the transparency and control of data- and model-driven decision making; (iv) Looking beyond the algorithm(s) for sources of bias and unfairness-in the myriad human decisions made during the problem formulation and modeling process; and (v) Supporting the cross-disciplinary scholarship necessary to do all of that well.

cs.CY↗

arXiv2016

Explaining Classification Models Built on High-Dimensional Sparse Data

Predictive modeling applications increasingly use data representing people's behavior, opinions, and interactions. Fine-grained behavior data often has different structure from traditional data, being very high-dimensional and sparse. Models built from these data are quite difficult to interpret, since they contain many thousands or even many millions of features. Listing features with large model coefficients is not sufficient, because the model coefficients do not incorporate information on feature presence, which is key when analysing sparse data. In this paper we introduce two alternatives for explaining predictive models by listing important features. We evaluate these alternatives in terms of explanation "bang for the buck,", i.e., how many examples' inferences are explained for a given number of features listed. The bottom line: (i) The proposed alternatives have double the bang-for-the-buck as compared to just listing the high-coefficient features, and (ii) interestingly, although they come from different sources and motivations, the two new alternatives provide strikingly similar rankings of important features.

stat.ML↗

arXiv2016

Enhancing Transparency and Control when Drawing Data-Driven Inferences about Individuals

Recent studies have shown that information disclosed on social network sites (such as Facebook) can be used to predict personal characteristics with surprisingly high accuracy. In this paper we examine a method to give online users transparency into why certain inferences are made about them by statistical models, and control to inhibit those inferences by hiding ("cloaking") certain personal information from inference. We use this method to examine whether such transparency and control would be a reasonable goal by assessing how difficult it would be for users to actually inhibit inferences. Applying the method to data from a large collection of real users on Facebook, we show that a user must cloak only a small portion of her Facebook Likes in order to inhibit inferences about their personal characteristics. However, we also show that in response a firm could change its modeling of users to make cloaking more difficult.

stat.ML↗

arXiv2006

Network-Based Marketing: Identifying Likely Adopters via Consumer Networks

Network-based marketing refers to a collection of marketing techniques that take advantage of links between consumers to increase sales. We concentrate on the consumer networks formed using direct interactions (e.g., communications) between consumers. We survey the diverse literature on such marketing with an emphasis on the statistical methods used and the data to which these methods have been applied. We also provide a discussion of challenges and opportunities for this burgeoning research topic. Our survey highlights a gap in the literature. Because of inadequate data, prior studies have not been able to provide direct, statistical support for the hypothesis that network linkage can directly affect product/service adoption. Using a new data set that represents the adoption of a new telecommunications service, we show very strong support for the hypothesis. Specifically, we show three main results: (1) ``Network neighbors''--those consumers linked to a prior customer--adopt the service at a rate 3--5 times greater than baseline groups selected by the best practices of the firm's marketing team. In addition, analyzing the network allows the firm to acquire new customers who otherwise would have fallen through the cracks, because they would not have been identified based on traditional attributes. (2) Statistical models, built with a very large amount of geographic, demographic and prior purchase data, are significantly and substantially improved by including network information. (3) More detailed network information allows the ranking of the network neighbors so as to permit the selection of small sets of individuals with very high probabilities of adoption.

math.ST↗

arXiv2000

Applications of Data Mining to Electronic Commerce

Electronic commerce is emerging as the killer domain for data mining technology. The following are five desiderata for success. Seldom are they they all present in one data mining application. 1. Data with rich descriptions. For example, wide customer records with many potentially useful fields allow data mining algorithms to search beyond obvious correlations. 2. A large volume of data. The large model spaces corresponding to rich data demand many training instances to build reliable models. 3. Controlled and reliable data collection. Manual data entry and integration from legacy systems both are notoriously problematic; fully automated collection is considerably better. 4. The ability to evaluate results. Substantial, demonstrable return on investment can be very convincing. 5. Ease of integration with existing processes. Even if pilot studies show potential benefit, deploying automated solutions to previously manual processes is rife with pitfalls. Building a system to take advantage of the mined knowledge can be a substantial undertaking. Furthermore, one often must deal with social and political issues involved in the automation of a previously manual business process.

cs.LG↗

arXiv2000

Robust Classification for Imprecise Environments

In real-world environments it usually is difficult to specify target operating conditions precisely, for example, target misclassification costs. This uncertainty makes building robust classification systems problematic. We show that it is possible to build a hybrid classifier that will perform at least as well as the best available classifier for any target conditions. In some cases, the performance of the hybrid actually can surpass that of the best known classifier. This robust performance extends across a wide variety of comparison frameworks, including the optimization of metrics such as accuracy, expected cost, lift, precision, recall, and workforce utilization. The hybrid also is efficient to build, to store, and to update. The hybrid is based on a method for the comparison of classifier performance that is robust to imprecise class distributions and misclassification costs. The ROC convex hull (ROCCH) method combines techniques from ROC analysis, decision analysis and computational geometry, and adapts them to the particulars of analyzing learned classifiers. The method is efficient and incremental, minimizes the management of classifier performance data, and allows for clear visual comparisons and sensitivity analyses. Finally, we point to empirical evidence that a robust hybrid classifier indeed is needed for many real-world problems.

cs.LG↗