arXiv subjects

Munther A. Dahleh

Publications and source records attributed to Munther A. Dahleh.

At least 19 recordsLinked to original sources

arXiv2026

Information Design with Elicitation and Strategic Coordination

We study linear-quadratic games of incomplete information with Gaussian uncertainty, where each player's payoff depends on a privately observed type and a common state. The designer observes the state, elicits types, and sells action recommendations. We characterize all implementable mechanisms with Gaussian joint distributions of actions and fundamentals, and identify the players-optimal, consumer-optimal, and revenue-maximizing designs. In games of strategic complements (substitutes), these optimal mechanisms maximally correlate (anticorrelate) players' actions. When type uncertainty is large, recommendations become deterministic linear functions of the state and reports, but remain only partially revealing.

cs.GT↗

arXiv2026

Modulated learning for private and distributed regression with just a single sample per client device

This work focuses on the question of learning from a large number of devices with each device holding only a single sample of data. Several real-world applications exist to this one sample per client setup up including learning from fitness trackers, data/app usage aggregators, body-worn sensing devices, and daily event monitors to name a few. When a client has only one sample, the standard federated learning paradigm breaks down as a local update based on that single point is far from being useful, especially in the earlier rounds for estimation of the model coefficients. This utility is further weakened by the privacy-inducing noise applied at every round. This work caters to this problem to enable such clients to collaboratively contribute to effectively learn a global model without leaking the privacy of their data. The proposed approach injects a single, carefully calibrated noisy perturbation to transform the sample at each client, followed by a post-processed representation which is shared with the server. These representations aggregated at the server are processed to obtain an unbiased gradient update that in expectation matches the non-private centralized gradient while preserving data privacy. This approach is different than traditional private federated learning, where the communication payloads involve model coefficients as opposed to privately transformed data samples. This method enables devices with extremely limited data to collaborate and learn accurate, privacy-preserving models without requiring large local datasets or sacrificing individual privacy.

cs.LG↗

arXiv2026

Compositional Online Learning for Multi-Objective System Co-Design

Many engineered systems must balance competing objectives, such as performance and safety, cost and reliability, or efficiency and sustainability, and are naturally modeled as compositions of interacting subsystems. We study online multi-objective decision-making in monotone co-design, where functionalities and resources are partially ordered, and the goal is to identify the target-feasible antichain of non-dominated trade-offs using few expensive evaluations. We introduce optimistic evaluators: history-dependent bounds on functionality and resource mappings that enable safe elimination of implementations before full evaluation. Based on these evaluators, we develop an elimination-based rejection-sampling algorithm, prove its soundness, and show that the admissible region shrinks monotonically as information accumulates. We instantiate the framework under monotonicity, Lipschitz continuity, and linear-parametric structure. For compositional co-design problems modeled by multigraphs, we show how local optimistic certificates propagate through the tractable remainder of the graph to yield system-level optimistic feasibility and resource bounds. Experiments on multi-robot fleet design, intermodal mobility systems, and synthetic monotone and Lipschitz benchmarks show substantial sample-efficiency gains over uniform sampling, Bayesian optimization, and multi-objective evolutionary algorithms.

math.OC↗

arXiv2025

Scheduling the Charge of Temporally Flexible Electric Vehicles: a Market-based Approach

The increasing electrification of human activities and the rapid integration of variable renewable energy sources strain the power grid. A solution to address the need for more grid storage is to use the battery of electric vehicles as a back-up capacity. However, drivers tend to disconnect their electric vehicle when its battery is needed the most. We propose a charge scheduler that incentivizes drivers to delay their disconnection to improve vehicle-to-grid services. We also leverage drivers' temporal flexibility to alleviate congestion in oversubscribed charging stations. We formulate the computation of an optimal flexible schedule as a mixed-integer quadratic problem. We tractably approximate its solution using the Alternating Direction Method of Multipliers. Considering the possibility that strategic drivers misreport their charging preferences to the station coordinator, we then propose a Vickrey-Clarke-Groves mechanism that incentivizes truthful reporting. We conclude with a simulated case study using real-world data to quantitatively assess the added value of drivers' temporal flexibility for enhancing vehicle-to-grid services and reducing station congestion.

eess.SY↗

arXiv2025

Eco-driving Incentive Mechanisms for Mitigating Emissions in Urban Transportation

This paper develops incentive mechanisms for promoting eco-driving with the overarching goal of minimizing emissions in transportation networks. The system operator provides drivers with energy-efficient driving guidance throughout their trips and measures compliance through vehicle telematics that capture how closely drivers follow this guidance. Drivers optimize their behaviors based on personal trade-offs between travel times and emissions. To design effective incentives, the operator elicits driver preferences regarding trip urgency and willingness to eco-drive, while determining optimal budget allocations and eco-driving recommendations. Two distinct settings based on driver behavior are analyzed. When drivers report their preferences truthfully, an incentive mechanism ensuring obedience (drivers find it optimal to follow recommendations) is designed by implementing eco-driving recommendations as a Nash equilibrium. When drivers may report strategically, the mechanism is extended to be both obedient and truthful (drivers find it optimal to report truthfully). Unlike existing works that focus on congestion or routing decisions in transportation networks, our framework explicitly targets emissions reduction by incentivizing drivers. The proposed mechanism addresses both strategic behavior and network effects arising from driver interactions, without requiring the operator to reveal system parameters to the drivers. Numerical simulations demonstrate the effects of budget constraints, driver types, and strategic misreporting on equilibrium outcomes and emissions reduction.

cs.GT↗

arXiv2025

Learning clusters of partially observed linear dynamical systems

We study the problem of learning clusters of partially observed linear dynamical systems from multiple input-output trajectories. This setting is particularly relevant when there are limited observations (e.g., short trajectories) from individual data sources, making direct estimation challenging. In such cases, incorporating data from multiple related sources can improve learning. We propose an estimation algorithm that leverages different data requirements for the tasks of clustering and system identification. First, short impulse responses are estimated from individual trajectories and clustered. Then, refined models for each cluster are jointly estimated using multiple trajectories. We establish end-to-end finite sample guarantees for estimating Markov parameters and state space realizations and highlight trade-offs among the number of observed systems, the trajectory lengths, and the complexity of the underlying models.

eess.SY↗

arXiv2025

Temperature-Controlled Smart Charging for Electric Vehicles in Cold Climates

The battery performance and lifespan of electric vehicles (EVs) degrade significantly in cold climates, requiring a considerable amount of energy to heat up the EV batteries. This paper proposes a novel technology, namely temperature-controlled smart charging, to coordinate the heating/charging power and reduce the total energy use of a solar-powered EV charging station. Instead of fixing the battery temperature setpoints, we analyze the thermal dynamics and inertia of EV batteries, and decide the optimal timing and proper amount of energy allocated for heating. In addition, a temperature-sensitive charging model is formulated with consideration of dynamic charging rates as well as battery health. We further tailor acceleration algorithms for large-scale EV charging, including the reduced-order dual decomposition and vehicle rescheduling. Simulation results demonstrate that the proposed temperature-controlled smart charging is superior in capturing the flexibility value of EV batteries and making full use of the rooftop solar energy. The proposed model typically achieves a 12.5--18.4% reduction in the charging cost and a 0.4--6.8% drop in the overhead energy use for heating.

eess.SY↗

arXiv2024

Mitigating Information Asymmetry in Two-Stage Contracts with Non-Myopic Agents

We consider a Stackelberg game in which a principal (she) establishes a two-stage contract with a non-myopic agent (he) whose type is unknown. The contract takes the form of an incentive function mapping the agent's first-stage action to his second-stage incentive. While the first-stage action reveals the agent's type under truthful play, a non-myopic agent could benefit from portraying a false type in the first stage to obtain a larger incentive in the second stage. The challenge is thus for the principal to design the incentive function so as to induce truthful play. We show that this is only possible with a constant, non-reactive incentive functions when the type space is continuous, whereas it can be achieved with reactive functions for discrete types. Additionally, we show that introducing an adjustment mechanism that penalizes inconsistent behavior across both stages allows the principal to design more flexible incentive functions.

cs.GT↗

arXiv2024

Incentive Design for Eco-driving in Urban Transportation Networks

Eco-driving emerges as a cost-effective and efficient strategy to mitigate greenhouse gas emissions in urban transportation networks. Acknowledging the persuasive influence of incentives in shaping driver behavior, this paper presents the `eco-planner,' a digital platform devised to promote eco-driving practices in urban transportation. At the outset of their trips, users provide the platform with their trip details and travel time preferences, enabling the eco-planner to formulate personalized eco-driving recommendations and corresponding incentives, while adhering to its budgetary constraints. Upon trip completion, incentives are transferred to users who comply with the recommendations and effectively reduce their emissions. By comparing our proposed incentive mechanism with a baseline scheme that offers uniform incentives to all users, we demonstrate that our approach achieves superior emission reductions and increased user compliance with a smaller budget.

eess.SY↗

arXiv2022

Data-driven control of COVID-19 in buildings: a reinforcement-learning approach

In addition to its public health crisis, COVID-19 pandemic has led to the shutdown and closure of workplaces with an estimated total cost of more than $16 trillion. Given the long hours an average person spends in buildings and indoor environments, this research article proposes data-driven control strategies to design optimal indoor airflow to minimize the exposure of occupants to viral pathogens in built environments. A general control framework is put forward for designing an optimal velocity field and proximal policy optimization, a reinforcement learning algorithm is employed to solve the control problem in a data-driven fashion. The same framework is used for optimal placement of disinfectants to neutralize the viral pathogens as an alternative to the airflow design when the latter is practically infeasible or hard to implement. We show, via simulation experiments, that the control agent learns the optimal policy in both scenarios within a reasonable time. The proposed data-driven control framework in this study will have significant societal and economic benefits by setting the foundation for an improved methodology in designing case-specific infection control guidelines that can be realized by affordable ventilation devices and disinfectants.

eess.SY↗

arXiv2022

Incentive Compatibility in Two-Stage Repeated Stochastic Games

We address the problem of mechanism design for two-stage repeated stochastic games -- a novel setting using which many emerging problems in next-generation electricity markets can be readily modeled. Repeated playing affords the players a large class of strategies that adapt a player's actions to all past observations and inferences obtained therefrom. In other settings such as iterative auctions or dynamic games where a large strategy space of this sort manifests, it typically has an important implication for mechanism design: It may be impossible to obtain truth-telling as a dominant strategy equilibrium. Consequently, in such scenarios, it is common to settle for mechanisms that render truth-telling only a Nash equilibrium, or variants thereof, even though Nash equilibria are known to be poor models of real-world behavior. This is owing to each player having to make overly specific assumptions about the behaviors of the other players to employ their Nash equilibrium strategy, which they may not make. In general, the lesser the burden of speculation in an equilibrium, the more plausible it is that it models real-world behavior. Guided by this maxim, we introduce a new notion of equilibrium called Dominant Strategy Non-Bankrupting Equilibrium (DNBE) which requires the players to make very little assumptions about the behavior of the other players to employ their equilibrium strategy. Consequently, a mechanism that renders truth-telling a DNBE as opposed to only a Nash equilibrium could be quite effective in molding real-world behavior along truthful lines. We present a mechanism for two-stage repeated stochastic games that renders truth-telling a Dominant Strategy Non-Bankrupting Equilibrium. The mechanism also guarantees individual rationality and maximizes social welfare. Finally, we describe an application of the mechanism to design demand response markets.

econ.TH↗

arXiv2022

A Two-Stage Mechanism for Demand Response Markets

Demand response involves system operators using incentives to modulate electricity consumption during peak hours or when faced with an incidental supply shortage. However, system operators typically have imperfect information about their customers' baselines, that is, their consumption had the incentive been absent. The standard approach to estimate the reduction in a customer's electricity consumption then is to estimate their counterfactual baseline. However, this approach is not robust to estimation errors or strategic exploitation by the customers and can potentially lead to overpayments to customers who do not reduce their consumption and underpayments to those who do. Moreover, optimal power consumption reductions of the customers depend on the costs that they incur for curtailing consumption, which in general are private knowledge of the customers, and which they could strategically misreport in an effort to improve their own utilities even if it deteriorates the overall system cost. The two-stage mechanism proposed in this paper circumvents the aforementioned issues. In the day-ahead market, the participating loads are required to submit only a probabilistic description of their next-day consumption and costs to the system operator for day-ahead planning. It is only in real-time, if and when called upon for demand response, that the loads are required to report their baselines and costs. They receive credits for reductions below their reported baselines. The mechanism for calculating the credits guarantees incentive compatibility of truthful reporting of the probability distribution in the day-ahead market and truthful reporting of the baseline and cost in real-time. The mechanism can be viewed as an extension of the celebrated Vickrey-Clarke-Groves mechanism augmented with a carefully crafted second-stage penalty for deviations from the day-ahead bids.

eess.SY↗

arXiv2022

Simple control for complex pandemics

The COVID-19 pandemic began over two years ago, yet schools, businesses, and other organizations are still struggling to keep the risk of disease outbreak low while returning to (near) normal functionality. Observations from these past years suggest that this goal can be achieved through the right balance of mitigation strategies, which may include some combination of mask use, vaccinations, viral testing, and contact tracing. The choice of mitigation measures will be uniquely based on the needs and available resources of each organization. This article presents practical guidance for creating these policies based on an analytical model of disease spread that captures the combined effects of each of these interventions. The resulting guidance is tested through simulation across a wide range of parameters and used to discuss the spread of disease on college campuses.

q-bio.PE↗

arXiv2021

Nonstochastic Bandits with Infinitely Many Experts

We study the problem of nonstochastic bandits with expert advice, extending the setting from finitely many experts to any countably infinite set: A learner aims to maximize the total reward by taking actions sequentially based on bandit feedback while benchmarking against a set of experts. We propose a variant of Exp4.P that, for finitely many experts, enables inference of correct expert rankings while preserving the order of the regret upper bound. We then incorporate the variant into a meta-algorithm that works on infinitely many experts. We prove a high-probability upper bound of $\tilde{\mathcal{O}} \big( i^*K + \sqrt{KT} \big)$ on the regret, up to polylog factors, where $i^*$ is the unknown position of the best expert, $K$ is the number of actions, and $T$ is the time horizon. We also provide an example of structured experts and discuss how to expedite learning in such case. Our meta-learning algorithm achieves optimal regret up to polylog factors when $i^* = \tilde{\mathcal{O}} \big( \sqrt{T/K} \big)$. If a prior distribution is assumed to exist for $i^*$, the probability of optimality increases with $T$, the rate of which can be fast.

cs.LG↗

arXiv2020

An Efficient and Incentive-Compatible Mechanism for Energy Storage Markets

A key obstacle to increasing renewable energy penetration in the power grid is the lack of utility-scale storage capacity. Transportation electrification has the potential to overcome this obstacle since Electric Vehicles (EVs) that are not in transit can provide battery storage as a service to the grid. This is referred to as EV-Power grid integration, and could potentially be a key milestone in the pathway to decarbonize the electricity and the transportation sectors. We first show that if EV-Power grid integration is not done carefully, then contrary to improving the cost efficiency of operating the grid, it could in fact be counterproductive to it. This fundamentally occurs due to two phenomena operating in tandem - the randomness of EV usage patterns and the possibility of strategic behavior by EV operators. We present a market-based solution to address this issue. Specifically, we develop a mechanism for energy storage markets using which the system operator can efficiently integrate a fleet of strategic EVs with random usage patterns into the grid, utilize them for storage, and satisfy the demand at minimum possible cost.

eess.SY↗

arXiv2020

A Cross-Domain Approach to Analyzing the Short-Run Impact of COVID-19 on the U.S. Electricity Sector

The novel coronavirus disease (COVID-19) has rapidly spread around the globe in 2020, with the U.S. becoming the epicenter of COVID-19 cases since late March. As the U.S. begins to gradually resume economic activity, it is imperative for policymakers and power system operators to take a scientific approach to understanding and predicting the impact on the electricity sector. Here, we release a first-of-its-kind cross-domain open-access data hub, integrating data from across all existing U.S. wholesale electricity markets with COVID-19 case, weather, cellular location, and satellite imaging data. Leveraging cross-domain insights from public health and mobility data, we uncover a significant reduction in electricity consumption across that is strongly correlated with the rise in the number of COVID-19 cases, degree of social distancing, and level of commercial activity.

cs.CY↗

arXiv2020

Data-driven control of micro-climate in buildings: an event-triggered reinforcement learning approach

Smart buildings have great potential for shaping an energy-efficient, sustainable, and more economic future for our planet as buildings account for approximately 40% of the global energy consumption. Future of the smart buildings lies in using sensory data for adaptive decision making and control that is currently gloomed by the key challenge of learning a good control policy in a short period of time in an online and continuing fashion. To tackle this challenge, an event-triggered -- as opposed to classic time-triggered -- paradigm, is proposed in which learning and control decisions are made when events occur and enough information is collected. Events are characterized by certain design conditions and they occur when the conditions are met, for instance, when a certain state threshold is reached. By systematically adjusting the time of learning and control decisions, the proposed framework can potentially reduce the variance in learning, and consequently, improve the control process. We formulate the micro-climate control problem based on semi-Markov decision processes that allow for variable-time state transitions and decision making. Using extended policy gradient theorems and temporal difference methods in a reinforcement learning set-up, we propose two learning algorithms for event-triggered control of micro-climate in buildings. We show the efficacy of our proposed approach via designing a smart learning thermostat that simultaneously optimizes energy consumption and occupants' comfort in a test building.

eess.SY↗

arXiv2020

Data Driven Estimation of Stochastic Switched Linear Systems of Unknown Order

We address the problem of learning the parameters of a mean square stable switched linear systems (SLS) with unknown latent space dimension, or \textit{order}, from its noisy input--output data. In particular, we focus on learning a good lower order approximation of the underlying model allowed by finite data. Motivated by subspace-based algorithms in system theory, we construct a Hankel-like matrix from finite noisy data using ordinary least squares. Such a formulation circumvents the non-convexities that arise in system identification, and allows for accurate estimation of the underlying SLS as data size increases. Since the model order is unknown, the key idea of our approach is model order selection based on purely data dependent quantities. We construct Hankel-like matrices from data of dimension obtained from the order selection procedure. By exploiting tools from theory of model reduction for SLS, we obtain suitable approximations via singular value decomposition (SVD) and show that the system parameter estimates are close to a balanced truncated realization of the underlying system with high probability.

eess.SY↗