This paper investigates ML systems serving a group of users, with multiple models/services, each aimed at specializing to a sub-group of users. We consider settings where upon deploying a set of services, users choose the one minimizing their personal losses and the learner iteratively learns by interacting with diverse users. Prior research shows that the outcomes of learning dynamics, which comprise both the services' adjustments and users' service selections, hinge significantly on the initial conditions. However, finding good initial conditions faces two main challenges: (i) \emph{Bandit feedback:} Typically, data on user preferences are not available before deploying services and observing user behavior; (ii) \emph{Suboptimal local solutions:} The total loss landscape (i.e., the sum of loss functions across all users and services) is not convex and gradient-based algorithms can get stuck in poor local minima. We address these challenges with a randomized algorithm to adaptively select a minimal set of users for data collection in order to initialize a set of services. Under mild assumptions on the loss functions, we prove that our initialization leads to a total loss within a factor of the \textit{globally optimal total loss,with complete user preference data}, and this factor scales logarithmically in the number of services. This result is a generalization of the well-known $k$-means++ guarantee to a broad problem class which is also of independent interest. The theory is complemented by experiments on real as well as semi-synthetic datasets.

Numerous online services are data-driven: the behavior of users affects the system's parameters, and the system's parameters affect the users' experience of the service, which in turn affects the way users may interact with the system. For example, people may choose to use a service only for tasks that already works well, or they may choose to switch to a different service. These adaptations influence the ability of a system to learn about a population of users and tasks in order to improve its performance broadly. In this work, we analyze a class of such dynamics -- where users allocate their participation amongst services to reduce the individual risk they experience, and services update their model parameters to reduce the service's risk on their current user population. We refer to these dynamics as \emph{risk-reducing}, which cover a broad class of common model updates including gradient descent and multiplicative weights. For this general class of dynamics, we show that asymptotically stable equilibria are always segmented, with sub-populations allocated to a single learner. Under mild assumptions, the utilitarian social optimum is a stable equilibrium. In contrast to previous work, which shows that repeated risk minimization can result in (Hashimoto et al., 2018; Miller et al., 2021), we find that repeated myopic updates with multiple learners lead to better outcomes. We illustrate the phenomena via a simulated example initialized from real data.

The impacts of link recommendations on social networks are challenging to evaluate, due to feedback loops between algorithmic recommendations and underlying network dynamics. Observational studies have limitations in answering causal questions; naive A/B experiments often result in biased evaluations due to unaccounted network interference and finally, existing simulations primarily employ static network models that do not take into account dynamics. Departing from existing approaches, we employ simulations to study dynamic impacts of link recommendations. Specifically, we propose an extension to the Jackson-Rogers network evolution model and investigate how link recommendations affect network evolution over time. Our experiments demonstrate that link recommendations can have surprising delayed and indirect effects on the structural properties of networks. Effects of recommendations vary in the short-term and long-term, such as the immediate reduction in degree inequality but eventual increase in degree inequality through friend-of-friend recommendations. Furthermore, even after recommendations are discontinued, their impacts can persist in the network, in part by altering natural network evolution dynamics. These results provide valuable insights into the interplay between algorithmic interventions and natural network dynamics and highlight the limitations of current evaluation paradigms.

Shape-Constrained Regression Using Algebraic Techniques How can one fit a multivariate polynomial to data points in such a way that this polynomial is guaranteed to have certain shape constraints, such as convexity or monotonicity in some of its variables? In “Shape-constrained regression using sum-of-squares polynomials,” M. Curmei and G. Hall propose a hierarchy of semidefinite programs to address this problem. They show that polynomial functions that are optimal to any fixed level of our hierarchy form a consistent estimator of the underlying shape-constrained function, which generates the data points. As a by-product of the proof, they establish that sum-of-squares-convex polynomials are dense in the set of polynomials that are convex over an arbitrary box. They further demonstrate the performance of the regressor for the problem of computing optimal transport maps in a color transfer task and that of estimating the optimal value function of a conic program. A real-time application of the latter problem to inventory management contract negotiation is presented.

We consider the problem of training private recommendation models with access to public item features. Training with Differential Privacy (DP) offers strong privacy guarantees, at the expense of loss in recommendation quality. We show that incorporating public item features during training can help mitigate this loss in quality. We propose a general approach based on collective matrix factorization (CMF), that works by simultaneously factorizing two matrices: the user feedback matrix (representing sensitive data) and an item feature matrix that encodes publicly available (non-sensitive) item information. The method is conceptually simple, easy to tune, and highly scalable. It can be applied to different types of public item data, including: (1) categorical item features; (2) item-item similarities learned from public sources; and (3) publicly available user feedback. Furthermore, these data modalities can be collectively utilized to fully leverage public data. Evaluating our method on a standard DP recommendation benchmark, we find that using public item features significantly narrows the quality gap between private models and their non-private counterparts. As privacy constraints become more stringent, models rely more heavily on public side features for recommendation. This results in a smooth transition from collaborative filtering to item-based contextual recommendations.

Designing recommendation systems that serve content aligned with time varying preferences requires proper accounting of the feedback effects of recommendations on human behavior and psychological condition. We argue that modeling the influence of recommendations on people's preferences must be grounded in psychologically plausible models. We contribute a methodology for developing grounded dynamic preference models. We demonstrate this method with models that capture three classic effects from the psychology literature: Mere-Exposure, Operant Conditioning, and Hedonic Adaptation. We conduct simulation-based studies to show that the psychological models manifest distinct behaviors that can inform system design. Our study has two direct implications for dynamic user modeling in recommendation systems. First, the methodology we outline is broadly applicable for psychologically grounding dynamic preference models. It allows us to critique recent contributions based on their limited discussion of psychological foundation and their implausible predictions. Second, we discuss implications of dynamic preference models for recommendation systems evaluation and design. In an example, we show that engagement and diversity metrics may be unable to capture desirable recommendation system performance.

Abstract Background With reduced community mobility, household infections may become increasingly important in SARS-CoV-2 transmission dynamics. Methods We investigate the intra-household transmission of COVID-19 through the secondary-attack rate (SAR) and household reproduction number (Rh). We estimate these using (i) data from 29 prior studies (February–August 2020), (ii) epidemiologically linked confirmed cases from Singapore (January–April 2020) and (iii) widespread-testing data from Vo’ (February–March 2020). For (i), we use a Bayesian random-effects model that corrects for reverse transcription–polymerase chain reaction (RT–PCR) test sensitivity and asymptomatic cases. We investigate the robustness of Rh with respect to community transmission rates and mobility patterns. Results The corrected pooled estimates from prior studies for SAR and Rh are 24% (20–28%) and 0.34 (0.30–0.38), respectively. Without corrections, the pooled estimates are: SAR = 18% (14–21%) and Rh = 0.28 (0.25–0.32). The corrected estimates line up with direct estimates from contact-tracing data from Singapore [Rh = 0.32 (0.22–0.42)] and population testing data from Vo’ [SAR = 31% (28–34%) and Rh = 0.37 (0.34–0.40)]. The analysis of Singapore data further suggests that the value of Rh (0.22–0.42) is robust to community-spread dynamics; our estimate of Rh stays constant whereas the fraction of infections attributable to household transmission (Rh/Reff) is lowest during outbreaks (5–7%) and highest during lockdowns and periods of low community spread (25–30%). Conclusions The three data-source types yield broadly consistent estimates for SAR and Rh. Our study suggests that household infections are responsible for a large fraction of infections and so household transmission may be an effective target for intervention.

In this work, we consider how preference models in interactive recommendation systems determine the availability of content and users' opportunities for discovery. We propose an evaluation procedure based on stochastic reachability to quantify the maximum probability of recommending a target piece of content to an user for a set of allowable strategic modifications. This framework allows us to compute an upper bound on the likelihood of recommendation with minimal assumptions about user behavior. Stochastic reachability can be used to detect biases in the availability of content and diagnose limitations in the opportunities for discovery granted to users. We show that this metric can be computed efficiently as a convex program for a variety of practical settings, and further argue that reachability is not inherently at odds with accuracy. We demonstrate evaluations of recommendation algorithms trained on large datasets of explicit and implicit ratings. Our results illustrate how preference models, selection rules, and user interventions impact reachability and how these effects can be distributed unevenly.

We study the problem of finding equilibrium strategies in multi-agent games with incomplete payoff information, where the payoff matrices are only known to the players up to some bounded uncertainty sets. In such games, an ex-post equilibrium characterizes equilibrium strategies that are robust to the payoff uncertainty. When the game is one-shot, we show that in zero-sum polymatrix games, an ex-post equilibrium can be computed efficiently using linear programming. We further extend the notion of ex-post equilibrium to stochastic games, where the game is played repeatedly in a sequence of stages and the transition dynamics are governed by an Markov decision process (MDP). We provide sufficient condition for the existence of an ex-post Markov perfect equilibrium (MPE). We show that under bounded payoff uncertainty, the value of any two-player zero-sum stochastic game can be computed up to a tight value interval using dynamic programming.

Access to digital content is often mediated by recommendations, which are primarily designed to predict user preferences. We take an alternate view in this work, exploring mechanisms which determine the availability of content from the perspective of user agency. We introduce a notion of reachability and show how it is determined by preference scoring functions, item selection rules, and user interface decisions. Simulated experiments illustrate these insights.

Introduction and Goals: SARS-CoV-2 is transmitted both in the community and within households. Social distancing and lockdowns reduce community transmission but do not directly address household transmission. We provide quantitative measures of household transmission based on empirical data, and estimate the contribution of households to overall spread. We highlight policy implications from our analysis of household transmission, and more generally, of changes in contact patterns under social distancing. Methods: We investigate the household secondary attack rate (SAR) for SARS-CoV-2, as well as R_h, which is the average number of within-household infections caused by a single index case. We estimate the SAR through a meta-analysis of previous studies. Taking a Bayesian approach, we correct estimates for the false-negative rate (FNR) of testing and compute posterior credible intervals for SAR. We estimate R_h using results from population testing in Vo', Italy and contact tracing data that we curate from Singapore. The code and data behind our analysis are publicly available at: https://github.com/andrewilyas/covid-household-transmission. Results: Our analysis of eight recent attack rate studies yields an SAR estimate of 0.28 (0.10, 0.61). From contact tracing data, we infer an R_h value of 0.32 (0.22, 0.42). Blanket testing data from Vo' yields an R_h estimate of 0.37 (0.34, 040) and a CRI estimate of 0.5 after correcting for FNR and asymptomatic cases. Interpretation: Our estimates of R_h suggest that household transmission was a small fraction (5%-35%) of R before social distancing but a large fraction after (30%-55%). This suggests that household transmission may be an effective target for interventions. A remaining uncertainty is whether household infections actually contribute to further community transmission or are contained within households. This can be estimated given high-quality contact tracing data. More broadly, our study points to emerging contact patterns (i.e., increased time at home relative to the community) playing a role in transmission of SARS-CoV-2. We briefly highlight another instance of this phenomenon (differences in contact between essential workers and the rest of the population), provide coarse estimates of its effect on transmission, and discuss how future data could enable a more reliable estimate.

Recommender systems operate in an inherently dynamical setting. Past recommendations influence future behavior, including which data points are observed and how user preferences change. However, experimenting in production systems with real user dynamics is often infeasible, and existing simulation-based approaches have limited scale. As a result, many state-of-the-art algorithms are designed to solve supervised learning problems, and progress is judged only by offline metrics. In this work we investigate the extent to which offline metrics predict online performance by evaluating eleven recommenders across six controlled simulated environments. We observe that offline metrics are correlated with online performance over a range of environments. However, improvements in offline metrics lead to diminishing returns in online performance. Furthermore, we observe that the ranking of recommenders varies depending on the amount of initial offline data available. We study the impact of adding exploration strategies, and observe that their effectiveness, when compared to greedy recommendation, is highly dependent on the recommendation algorithm. We provide the environments and recommenders described in this paper as Reclab: an extensible ready-to-use simulation framework at this https URL.

Since the mid-90s there has been a widely-held belief that signature files are inferior to inverted files for text indexing. In recent years the Bing search engine has developed and deployed an index based on bit-sliced signatures. This index, known as BitFunnel, replaced an existing production system based on an inverted index. The driving factor behind the shift away from the inverted index was operational cost savings. This paper describes algorithmic innovations and changes in the cloud computing landscape that led us to reconsider and eventually field a technology that was once considered unusable. The BitFunnel algorithm directly addresses four fundamental limitations in bit-sliced block signatures. At the same time, our mapping of the algorithm onto a cluster offers opportunities to avoid other costs associated with signatures. We show these innovations yield a significant efficiency gain versus classic bit-sliced signatures and then compare BitFunnel with Partitioned Elias-Fano Indexes, MG4J, and Lucene.