arXiv subjects

Van-Anh Truong

Publications and source records attributed to Van-Anh Truong.

8 recordsLinked to original sources

arXiv2021

Online Learning of Independent Cascade Models with Node-level Feedback

We propose a detailed analysis of the online-learning problem for Independent Cascade (IC) models under node-level feedback. These models have widespread applications in modern social networks. Existing works for IC models have only shed light on edge-level feedback models, where the agent knows the explicit outcome of every observed edge. Little is known about node-level feedback models, where only combined outcomes for sets of edges are observed; in other words, the realization of each edge is censored. This censored information, together with the nonlinear form of the aggregated influence probability, make both parameter estimation and algorithm design challenging. We establish the first confidence-region result under this setting. We also develop an online algorithm achieving a cumulative regret of $\mathcal{O}( \sqrt{T})$, matching the theoretical regret bound for IC models with edge-level feedback.

cs.SI↗

arXiv2021

Online Learning and Optimization Under a New Linear-Threshold Model with Negative Influence

Problem definition: Corporate brands, grassroots activists, and ordinary citizens all routinely employ Word-of-mouth (WoM) diffusion to promote products and instigate social change. Our work models the formation and spread of negative attitudes via WoM on a social network represented by a directed graph. In an online learning setting, we examine how an agent could simultaneously learn diffusion parameters and choose sets of seed users to initiate diffusions and maximize positive influence. In contrast to edge-level feedback, in which an agent observes the relationship (edge) through which a user (node) is influenced, we more realistically assume node-level feedback, where an agent only observes when a user is influenced and whether that influence is positive or negative. Methodology/results: We propose a new class of negativity-aware Linear Threshold Models. We show that in these models, the expected positive influence spread is a monotone submodular function of the seed set. Therefore, when maximizing positive influence by selecting a seed set of fixed size, a greedy algorithm can guarantee a solution with a constant approximation ratio. For the online learning setting, we propose an algorithm that runs in epochs of growing lengths, each consisting of a fixed number of exploration rounds followed by an increasing number of exploitation rounds controlled by a hyperparameter. Under mild assumptions, we show that our algorithm achieves asymptotic expected average scaled regret that is inversely related to any fractional constant power of the number of rounds. Managerial implications: During seed selection, our negativity-aware models and algorithms allow WoM campaigns to discover and best account for characteristics of local users and propagated content. We also give the first algorithms with regret guarantees for influence maximization under node-level feedback.

cs.SI↗

arXiv2020

Online Learning with Cumulative Oversampling: Application to Budgeted Influence Maximization

We propose a cumulative oversampling (CO) method for online learning. Our key idea is to sample parameter estimations from the updated belief space once in each round (similar to Thompson Sampling), and utilize the cumulative samples up to the current round to construct optimistic parameter estimations that asymptotically concentrate around the true parameters as tighter upper confidence bounds compared to the ones constructed with standard UCB methods. We apply CO to a novel budgeted variant of the Influence Maximization (IM) semi-bandits with linear generalization of edge weights, whose offline problem is NP-hard. Combining CO with the oracle we design for the offline problem, our online learning algorithm simultaneously tackles budget allocation, parameter learning, and reward maximization. We show that for IM semi-bandits, our CO-based algorithm achieves a scaled regret comparable to that of the UCB-based algorithms in theory, and performs on par with Thompson Sampling in numerical experiments.

cs.LG↗

arXiv2019

Beyond Adaptive Submodularity: Adaptive Influence Maximization with Intermediary Constraints

We consider a brand with a given budget that wants to promote a product over multiple rounds of influencer marketing. In each round, it commissions an influencer to promote the product over a social network, and then observes the subsequent diffusion of the product before adaptively choosing the next influencer to commission. This process terminates when the budget is exhausted. We assume that the diffusion process follows the popular Independent Cascade model. We also consider an online learning setting, where the brand initially does not know the diffusion parameters associated with the model, and has to gradually learn the parameters over time. Unlike in existing models, the rounds in our model are correlated through an intermediary constraint: each user can be commissioned for an unlimited number of times. However, each user will spread influence without commission at most once. Due to this added constraint, the order in which the influencers are chosen can change the influence spread, making obsolete existing analysis techniques that based on the notion of adaptive submodularity. We devise a sample path analysis to prove that a greedy policy that knows the diffusion parameters achieves at least $1-1/e - ε$ times the expected reward of the optimal policy. In the online-learning setting, we are the first to consider a truly adaptive decision making framework, rather than assuming independent epochs, and adaptivity only within epochs. Under mild assumptions, we derive a regret bound for our algorithm. In our numerical experiments, we simulate information diffusions on four Twitter sub-networks, and compare our UCB-based learning algorithms with several baseline adaptive seeding strategies. Our learning algorithm consistently outperforms the baselines and achieves rewards close to the greedy policy that knows the true diffusion parameters.

cs.LG↗

arXiv2019

Prophet Inequality with Correlated Arrival Probabilities, with Application to Two Sided Matchings

The classical Prophet Inequality arises from a fundamental problem in optimal-stopping theory. In this problem, a gambler sees a finite sequence of independent, non-negative random variables. If he stops the sequence at any time, he collects a reward equal to the most recent observation. The Prophet Inequality states that, knowing the distribution of each random variable, the gambler can achieve at least half as much reward in expectation, as a prophet who knows the entire sample path of random variables (Krengel and Sucheston 1978). In this paper, we prove a corresponding bound for correlated non-negative random variables. We analyze two methods for proving the bound, a constructive approach, which produces a worst-case instance, and a reductive approach, which characterizes a certain submartingale arising from the reward process of our online algorithm. We apply this new prophet inequality to the design of algorithms for a class of two-sided bipartite matching problems that underlie online task assignment problems. In these problems, demand units of various types arrive randomly and sequentially over time according to some stochastic process. Tasks, or supply units, arrive according to another stochastic process. Each demand unit must be irrevocably matched to a supply unit or rejected. The match earns a reward that depends on the pair. The objective is to maximize the total expected reward over the planning horizon. The problem arises in mobile crowd-sensing and crowd sourcing contexts, where workers and tasks must be matched by a platform according to various criteria. We derive the first online algorithms with worst-case performance guarantees for our class of two-sided bipartite matching problems.

math.OC↗

arXiv2018

Online Advance Admission Scheduling for Services with Customer Preferences

We study web and mobile applications that are used to schedule advance service, from medical appointments to restaurant reservations. We model them as online weighted bipartite matching problems with non-stationary arrivals. We propose new algorithms with performance guarantees for this class of problems. Specifically, we show that the expected performance of our algorithms is bounded below by $1-\sqrt{\frac{2}π}\frac{1}{\sqrt{k}}+O(\frac{1}{k})$ times that of an optimal offline algorithm, which knows all future information upfront, where $k$ is the minimum capacity of a resource. This is the tightest known lower bound. This performance analysis holds for any Poisson arrival process. Our algorithms can also be applied to a number of related problems, including display ad allocation problems and revenue management problems for opaque products. We test the empirical performance of our algorithms against several well-known heuristics by using appointment scheduling data from a major academic hospital system in New York City. The results show that the algorithms exhibit the best performance among all the tested policies. In particular, our algorithms are $21\%$ more effective than the actual scheduling strategy used in the hospital system according to our performance metric.

math.OC↗

arXiv2018

Advance Service Reservations with Heterogeneous Customers

We study a fundamental model of resource allocation in which a finite number of resources must be assigned in an online manner to a heterogeneous stream of customers. The customers arrive randomly over time according to known stochastic processes. Each customer requires a specific amount of capacity and has a specific preference for each of the resources, with some resources being feasible for the customer and some not. The system must find a feasible assignment of each customer to a resource or must reject the customer. The aim is to maximize the total expected capacity utilization of the resources over the horizon. This model has application in services, freight transportation, and online advertising. We present online algorithms with bounded competitive ratios relative to an optimal offline algorithm that knows all stochastic information. Our algorithms perform extremely well compared to common heuristics, as demonstrated on a real data set from a large hospital system in New York City.

math.OC↗

arXiv2015

Online Resource Allocation with Customer Choice

We introduce a general model of resource allocation with customer choice. In this model, there are multiple resources that are available over a finite horizon. The resources are non-replenishable and perishable. Each unit of a resource can be instantly made into one of several products. There are multiple customer types arriving randomly over time. An assortment of products must be offered to each arriving customer, depending on the type of the customer, the time of arrival, and the remaining inventory. From this assortment, the customer selects a product according to a general choice model. The selection generates a product-dependent and customer-type-dependent reward. The objective of the system is to maximize the total expected reward earned over the horizon. The above problem has a number of applications, including personalized assortment optimization, revenue management of parallel flights, and web- and mobile-based appointment scheduling. We derive online algorithms that are asymptotically optimal and achieve the best constant relative performance guarantees for this class of problems.

math.OC↗