In online platforms like eBay, sponsored search advertising has become instrumental for businesses aiming for enhanced visibility. However, in automated ad auctions, the sellers (ad campaigns) run the risk of exhausting their budgets prematurely in the absence of proper pacing strategies. In response to this, online platforms have been prompted to employ budget pacing strategies to maintain consistent spending patterns for their sellers. While numerous budget pacing strategies have been introduced, they predominantly stem from either empirical or theoretical perspectives, often functioning in isolation. This paper aims to bridge this gap by investigating the performance of a theoretically inspired optimization-based bid shading method, AdaptivePacing, within eBay's sponsored search environment and proposing variants of the algorithm tailored to real-world environments. Our findings highlight the benefits of applying theoretical pacing approaches in practical contexts. Specifically, the optimization-based AdaptivePacing method offers the platform flexible control over campaign spending patterns, accounts for business constraints, and suggests tailored strategies for distinct advertisers. Furthermore, when evaluating AdaptivePacing alongside established empirical methods, we demonstrate its practical effectiveness and pinpoint areas for further refinement.

Today's online platforms heavily lean on algorithmic recommendations for bolstering user engagement and driving revenue. However, these recommendations can impact multiple stakeholders simultaneously---the platform, items (sellers), and users (customers)---each with their unique objectives, making it difficult to find the right middle ground that accommodates all stakeholders. To address this, we introduce a novel fair recommendation framework, Problem (FAIR), that flexibly balances multi-stakeholder interests via a constrained optimization formulation. We next explore Problem (FAIR) in a dynamic online setting where data uncertainty further adds complexity, and propose a low-regret algorithm FORM that concurrently performs real-time learning and fair recommendations, two tasks that are often at odds. Via both theoretical analysis and a numerical case study on real-world data, we demonstrate the efficacy of our framework and method in maintaining platform revenue while ensuring desired levels of fairness for both items and users.

Traditional multi-armed bandit (MAB) frameworks, predominantly examined under stochastic or adversarial settings, often overlook the temporal dynamics inherent in many real-world applications such as recommendation systems and online advertising. This paper introduces a novel non-stationary MAB framework that captures the temporal structure of these real-world dynamics through an auto-regressive (AR) reward structure. We propose an algorithm that integrates two key mechanisms: (i) an alternation mechanism adept at leveraging temporal dependencies to dynamically balance exploration and exploitation, and (ii) a restarting mechanism designed to discard out-of-date information. Our algorithm achieves a regret upper bound that nearly matches the lower bound, with regret measured against a robust dynamic benchmark. Finally, via a real-world case study on tourism demand prediction, we demonstrate both the efficacy of our algorithm and the broader applicability of our techniques to more complex, rapidly evolving time series.

In the study of infectious diseases on networks, researchers calculate epidemic thresholds to help forecast whether or not a disease will eventually infect a large fraction of a population. Because network structure typically changes with time, which fundamentally influences the dynamics of spreading processes and in turn affects epidemic thresholds for disease propagation, it is important to examine epidemic thresholds in models of disease spread on temporal networks. Most existing studies of epidemic thresholds in temporal networks have focused on models in discrete time, but most real-world networked systems evolve continuously with time. In our work, we encode the continuous time-dependence of networks in the evaluation of the epidemic threshold of a susceptible–infected–susceptible (SIS) process by studying an SIS model on tie-decay networks. We derive the epidemic-threshold condition of this model, and we perform numerical experiments to verify it. We also examine how different factors—the decay coefficients of the tie strengths in a network, the frequency of the interactions between the nodes in the network, and the sparsity of the underlying social network on which interactions occur—lead to decreases or increases of the critical values of the threshold and hence contribute to facilitating or impeding the spread of a disease. We thereby demonstrate how the features of tie-decay networks alter the outcome of disease spread.

Many online platforms, ranging from online retail stores to social media platforms, employ algorithms to optimize their offered assortment of items (e.g., products and contents). These algorithms often focus exclusively on achieving the platforms' objectives, highlighting items with the highest popularity or revenue. This approach, however, can compromise the equality of opportunities for the rest of the items, in turn leading to less content diversity and increased regulatory scrutiny for the platform. Motivated by that, we introduce and study a fair assortment planning problem, which requires any two items with similar quality/merits to be offered similar outcomes. We show that the problem can be formulated as a linear program (LP), called (FAIR), that optimizes over the distribution of all feasible assortments. To find a near-optimal solution to (FAIR), we propose a framework based on the Ellipsoid method, which requires a polynomial-time separation oracle to the dual of the LP. We show that finding an optimal separation oracle to the dual problem is an NP-complete problem, and hence we propose a series of approximate separation oracles, which then result in a 1/2-approx. algorithm and an FPTAS for the original Problem (FAIR). The approximate separation oracles are designed by (i) showing the separation oracle to the dual of the LP is equivalent to solving an infinite series of parameterized knapsack problems, and (ii) leveraging the structure of the parameterized knapsack problems. Finally, we conduct a case study using the MovieLens dataset, which demonstrates the efficacy of our algorithms and further sheds light on the price of fairness.

In this work, we study a dynamic multi-armed bandit (MAB) problem, where the expected reward of each arm evolves over time following an auto-regressive model. We present an algorithm whose per-round regret upper bound almost matches the regret lower bound, and numerically demonstrate its efficacy in adapting to the changing environment.

An active area of research in computational science is the design of algorithms for solving the subgraph matching problem to find copies of a given template graph in a larger world graph. Prior works have largely addressed single-channel networks using a variety of approaches. We present a suite of filtering methods for subgraph isomorphisms for multiplex networks (with different types of edges between nodes and more than one edge within each channel type). We aim to understand the entire solution space rather than focusing on finding one isomorphism. Results are shown on several classes of datasets: (a) Sudoku puzzles mapped to the subgraph isomorphism problem, (b) Erdos-Renyi multigraphs, (c) real-world datasets from Twitter and transportation networks, (d) synthetic data created for the DARPA MAA program.

Multi-armed bandit (MAB) problems are mainly studied under two extreme settings known as stochastic and adversarial. These two settings, however, do not capture realistic environments such as search engines and marketing and advertising, in which rewards stochastically change in time. Motivated by that, we introduce and study a dynamic MAB problem with stochastic temporal structure, where the expected reward of each arm is governed by an auto-regressive (AR) model. Due to the dynamic nature of the rewards, simple "explore and commit" policies fail, as all arms have to be explored continuously over time. We formalize this by characterizing a per-round regret lower bound, where the regret is measured against a strong (dynamic) benchmark. We then present an algorithm whose per-round regret almost matches our regret lower bound. Our algorithm relies on two mechanisms: (i) alternating between recently pulled arms and unpulled arms with potential, and (ii) restarting. These mechanisms enable the algorithm to dynamically adapt to changes and discard irrelevant past information at a suitable rate. In numerical studies, we further demonstrate the strength of our algorithm under different types of non-stationary settings.

We present an approach for inexact subgraph matching on attributed graphs optimizing the graph edit distance. By combining lower bounds on the cost of individual assignments, we obtain a heuristic for a backtracking tree search to identify optimal solutions. We evaluate our algorithm on a knowledge graph dataset derived from real-world data, and analyze the space of optimal solutions.

In order for an e-commerce platform to maximize its revenue, it must recommend customers items they are most likely to purchase.However, the company often has business constraints on these items, such as the number of each item in stock.In this work, our goal is to recommend items to users as they arrive on a webpage sequentially, in an online manner, in order to maximize reward for a company, but also satisfy budget constraints.We first approach the simpler online problem in which the customers arrive as a stationary Poisson process, and present an integrated algorithm that performs online optimization and online learning together.We then make the model more complicated but more realistic, treating the arrival processes as non-stationary Poisson processes.To deal with heterogeneous customer arrivals, we propose a time segmentation algorithm that converts a non-stationary problem into a series of stationary problems.Experiments conducted on large-scale synthetic data demonstrate the effectiveness and efficiency of our proposed approaches on solving constrained resource allocation problems.

We present filtering methods for finding all sub-graphs of a large multiplex network that are isomorphic to a smaller template network. These methods are shown to be effective on a set of synthetic transaction networks from the DARPA Modeling Adversarial Activity (MAA) program. In some cases, filtering allows us to identify and enumerate all possible isomorphisms. We observe that in some of the MAA networks, the number of subgraphs isomorphic to the template is orders of magnitude larger than the size of the network.