Strategic Agents Research Articles

Approximate and strategyproof maximin share allocation of chores with ordinal preferences

We initiate the work on maximin share (MMS) fair allocation of m indivisible chores to n agents using only their ordinal preferences, from both algorithmic and mechanism design perspectives. The previous best-known approximation is 2-1/n by Aziz et al. [IJCAI 2017]. We improve this result by giving a simple deterministic 5/3-approximation algorithm that determines an allocation sequence of agents, according to which items are allocated one by one. By a tighter analysis, we show that for n=2,3, our algorithm achieves better approximation ratios, and is actually optimal. We also consider the setting with strategic agents, where agents may misreport their preferences to manipulate the outcome. We first provide a O(\log (m/n))-approximation consecutive picking algorithm, and then improve the approximation ratio to O(\sqrt{\log n}) by a randomized algorithm. Our results uncover some interesting contrasts between the approximation ratios achieved for chores versus goods.

Matching strategic agents on a two-sided platform

A platform offers sellers and buyers trading opportunities by creating one-to-one matches between them. A matching mechanism specifies subscription fees and the probabilities with which each seller type is matched with each buyer type. When the subscribers are fully strategic in their interactions with their matched partners, the optimal mechanism may not employ socially efficient positive assortative matching (PAM) but instead focus on the extraction of the agents' informational rents. In alternative scenarios in which the platform exercises stronger control over the subscribers' interactions, the optimal mechanism employs PAM but may create distortions by blocking some efficient transactions.

Truthful and Fair Mechanisms for Matroid-Rank Valuations

We study the problem of allocating indivisible goods among strategic agents. We focus on settings wherein monetary transfers are not available and each agent's private valuation is a submodular function with binary marginals, i.e., the agents' valuations are matroid-rank functions. In this setup, we establish a notable dichotomy between two of the most well-studied fairness notions in discrete fair division; specifically, between envy-freeness up to one good (EF1) and maximin shares (MMS). First, we show that a known Pareto-efficient mechanism is group strategy-proof for finding EF1 allocations, under matroid-rank valuations. The group strategy-proofness guarantee strengthens an existing result that establishes truthfulness (individually for each agent) in the same context. Our result also generalizes prior work from binary additive valuations to the matroid-rank case. Next, we establish that an analogous positive result cannot be achieved for MMS, even when considering truthfulness on an individual level. Specifically, we prove that, for matroid-rank valuations, there does not exist a truthful mechanism that is index oblivious, Pareto efficient, and maximin fair. For establishing our results, we develop a characterization of truthful mechanisms for matroid-rank functions. This characterization in fact holds for a broader class of valuations (specifically, holds for binary XOS functions) and might be of independent interest.

Sequential Blocked Matching

We consider a sequential blocked matching (SBM) model where strategic agents repeatedly report ordinal preferences over a set of services to a central planner. The planner's goal is to elicit agents' true preferences and design a policy that matches services to agents in order to maximize the expected social welfare with the added constraint that each matched service can be blocked or unavailable for a number of time periods. Naturally, SBM models the repeated allocation of reusable services to a set of agents where each allocated service becomes unavailable for a fixed duration. We first consider the offline SBM setting, where the strategic agents are aware of their true preferences. We measure the performance of any policy by distortion, the worst-case multiplicative approximation guaranteed by any policy. For the setting with s services, we establish lower bounds of Ω(s) and Ω(√s) on the distortions of any deterministic and randomised mechanisms, respectively. We complement these results by providing approximately truthful, measured by incentive ratio, deterministic and randomised policies based on random serial dictatorship which match our lower bounds. Our results show that there is a significant improvement if one considers the class of randomised policies. Finally, we consider the online SBM setting with bandit feedback where each agent is initially unaware of her true preferences, and the planner must facilitate each agent in the learning of their preferences through the matching of services over time. We design an approximately truthful mechanism based on the explore-then-commit paradigm, which achieves logarithmic dynamic approximate regret.

Truthful Aggregation of Budget Proposals with Proportionality Guarantees

We study a participatory budgeting problem, where a set of strategic agents wish to split a divisible budget among different projects by aggregating their proposals on a single division. Unfortunately, the straightforward rule that divides the budget proportionally is susceptible to manipulation. Recently, a class of truthful mechanisms has been proposed, namely the moving phantom mechanisms. One such mechanism satisfies the proportionality property, in the sense that in the extreme case where all agents prefer a single project to receive the whole amount, the budget is assigned proportionally. While proportionality is a naturally desired property, it is defined over a limited type of preference profiles. To address this, we expand the notion of proportionality, by proposing a quantitative framework that evaluates a budget aggregation mechanism according to its worst-case distance from the proportional allocation. Crucially, this is defined for every preference profile. We study this measure on the class of moving phantom mechanisms, and we provide approximation guarantees. For two projects, we show that the Uniform Phantom mechanism is optimal among all truthful mechanisms. For three projects, we propose a new, proportional mechanism that is optimal among all moving phantom mechanisms. Finally, we provide impossibility results regarding the approximability of moving phantom mechanisms.

Truthful Cake Sharing

The classic cake cutting problem concerns the fair allocation of a heterogeneous resource among interested agents. In this paper, we study a public goods variant of the problem, where instead of competing with one another for the cake, the agents all share the same subset of the cake which must be chosen subject to a length constraint. We focus on the design of truthful and fair mechanisms in the presence of strategic agents who have piecewise uniform utilities over the cake. On the one hand, we show that the leximin solution is truthful and moreover maximizes an egalitarian welfare measure among all truthful and position oblivious mechanisms. On the other hand, we demonstrate that the maximum Nash welfare solution is truthful for two agents but not in general. Our results assume that mechanisms can block each agent from accessing parts that the agent does not claim to desire; we provide an impossibility result when blocking is not allowed.

Strategic stiffening/cooling in the Ising game

The dynamic noisy binary choice (Ising) game of forward-looking agents on a complete graph is analysed. It is shown that strategic considerations lead to effective interaction strengthening (noise reduction) as compared to the myopic game. We show that strategic agents are able to come to consensus in the wider range of noise values than myopic ones. Effective population dynamics with time-dependent probabilities reflecting this strategic stiffening/cooling effect is described.

Supply–demand equilibria and multivariate optimization of social welfare in double-ended queueing systems

This paper considers a double-ended queueing system in which both the supply and demand sides adopt strategic behaviors in observable and unobservable cases. Systems with only one type of strategic agents are common in literature. Agents in these systems follow the crowd and form a common strategy adopted by all at the evolutionary endpoint, which is usually referred to as the “equilibrium strategy”. When both sides are strategic, however, the strategy adopted by any individual on one side depends not only on what the individuals of their own side are doing but also on the other side, which forms a Nash equilibrium between the two sides. To solve for such an equilibrium, we extend the existing game theoretical framework on queueing problems under the scope of multi-population games. In contrast, in the scenario of sellers and buyers being socially concerned, a joint social welfare function needs a multivariate maximization on the two decision variables. We solve multivariate optimization problems in both observable and unobservable cases to find socially optimal strategies for both sides.

Comparative analysis of risk-aversion and strategic investment behavior in generation expansion

Both risk-averse behavior and strategic bidding may distort the outcome of energy and capacity markets. Although these two behavioral aspects may trigger similar bidding behavior and investments into generation capacity, they are fundamentally different.While in the literature risk-aversion and strategic behavior have been investigated separately, the presented research offers a comparative analysis to identify situations in which both types of behavior result in the same outcome. First, a framework is introduced in which both strategic and risk-averse behavior can be investigated for a generation expansion planning problem. The problem including the risk-averse investor is modeled as a Nash-equilibrium problem and solved via Alternating Direction Method of Multipliers, whereas the Stackelberg game between the strategic agent and the markets is formulated as a bi-level optimization problem and solved using disjunctive constraints.Using a set of 1920 problem instances we reveal that the distinguishability of the two types of behavior is highly technology-dependent and influenced by the way the capacity target is set. Implications of this work may support regulators in disentangling risk-aversion and strategic behavior based on observed market outcomes.

Strategic bidding of wind power producers in electricity markets in presence of information sharing

In order to facilitate the growing share of Renewable Energy Sources (RES) in power systems around the world, a cost-effective integration of RES into the electricity market is essential. In addition to numerous technical challenges, the integration of RES-based generation creates some market design challenges. For example, monitoring market power abuse by RES-based electricity generation companies may be not straightforward, given the variable and limitedly predictable availability of these resources, not known by the regulator. In this context, this paper investigates the impact of sharing wind power forecasts among market participants on wind power producers’ (WPPs’) bidding behaviour. An analytical model that captures the strategic price-making behaviour of competing WPPs on the day-ahead electricity market is provided. A best response function is provided, showing the strategic quantity bid of a price-making WPP in function of the bids of other market participants. Subsequently, a ‘multiple leaders-common market clearing’ problem is formulated as a Nash game between the strategic WPPs. Last, we allow WPPs to make their individual wind power forecasts available to their competing WPPs. When all price-making market participants possess the forecasts of each WPP, all strategic agents estimate the market outcome correctly. On the contrary, with minimal information sharing, the price-making WPPs use an assumption on the behaviour of the other WPPs in the strategic determination of their quantity bids. Sharing information on individual forecasts may lead to a higher or lower wind power participation into the day-ahead market, profits for WPPs and social welfare. On the one hand, systems in which all the WPPs would bid above their mean forecast lead to a lower wind power participation when sharing individual forecasts among WPPs, hence, a lower social welfare. Additionally, the profits for WPPs increase. On the other hand, a higher wind power participation is realised when sharing forecasts for systems in which all WPPs bid below their mean forecast, hence a higher social welfare and lower WPP profits. These insights are useful for policy makers and regulators in order to ensure transparent and competitive electricity markets.

Strategi Komunikasi Pemasaran Agen PT Prudential Life Assurance Cabang Binjai


 Marketing Communication Strategy Agent PT. Prudential life assurance binjai branch develops effective marketing communications, namely identifying the target audience, designing messages, selecting media, selecting message sources, collecting feedback. Matthew B. Miles and A. Michael Huberman developed a qualitative descriptive method using data model interactivity analysis techniques. The results of the research obtained an overview that PT. Prudential in marketing strategic communication Agents are more than two-way communication to customers to sell products so that they can give influence, wealth, and no one is judged.
 Keywords: strategy, marketing communication, decision making, customer.
 
 

Auction and Classification of Smart Contracts

The execution of smart contracts (SCs) relies on consensus algorithms that validate the miner who executes the contract and gets a fee to cover her expenditure. In this sense, miners are strategic agents who may focus on executing those contracts with the largest fee, to the detriment of other SCs’ execution times, which also harms the blockchain’s reputation. This paper analyzes the impact of miners’ competition on SCs’ execution times in a public blockchain. First, we explain that the Proof-of-Work mechanism casts similarities with a time auction, where the one who first adds blocks is the one who executes the contract and gets the fee. At equilibrium, costs negatively affect execution times, while the opposite holds concerning fees. However, this result does not capture the competition for other contracts; hence, we apply the Naïve Bayes method to classify SCs by considering a simulated database that comprises miners’ competition for several contracts. We observe that simultaneous competition generates patterns that differ from the ones expected by the auction solution. For example, miners’ valuation does not accelerate contracts’ execution, and high-cost smart contracts do not necessarily execute at last places.

A Framework for Studying Decentralized Bayesian Learning with Strategic Agents

We study the problem of Bayesian learning in a dynamical system involving strategic agents with asymmetric information. In a series of seminal papers in the literature, this problem has been investigated under a simplifying model where selfish players appear sequentially and act once in the game. It has been shown that there exist information cascades where users discard their private information and mimic the action of their predecessor. In this paper, we provide a framework for studying Bayesian learning dynamics in a more general setting than the one just described. In particular, our model incorporates cases where players can act repeatedly and there is strategic interaction in that each agent’s payoff may also depend on other players’ actions. The proposed framework hinges on a sequential decomposition methodology for finding structured perfect Bayesian equilibria of a general class of dynamic games with asymmetric information. Using this methodology, we study a specific dynamic learning model where players make decisions about public investment based on their estimates of everyone’s states. We characterize a set of informational cascades for this problem where learning stops for the team as a whole. Moreover, we show that such cascades occur almost surely.

Pooling Agents for Customer-Intensive Services

To Pool or Not to Pool? Analyzing Customer-Intensive Services with Strategic Agents In customer-intensive services where service quality increases with service time, service providers commonly pool their agents and give performance bonuses that reward agents for achieving greater customer satisfaction and serving more customers. Conventional wisdom suggests that pooling agents reduce customer wait time whereas performance bonuses motivate agents to produce high-quality services, both of which should boost customer satisfaction. However, in “Pooling Agents for Customer-Intensive Services,” Wang, Yang, Cui, Ülkü, and Zhou find that when agents act strategically, they may choose to speed up under pooling in an attempt to serve more customers, thus undermining service quality. If this happens, pooling can backfire and result in both lower customer satisfaction and agent payoff. Consequently, the researchers propose a simple practical solution to restore the efficiency of pooling. They propose pooling a portion of the performance bonuses (incentive pooling) in conjunction with pooling agents (operational pooling).

Distributed dynamics for aggregative games: Robustness and privacy guarantees

AbstractThis paper considers the problem of Nash equilibrium (NE) seeking in aggregative games, where the cost function of each player depends on an aggregate of all players' actions. We present a distributed continuous‐time algorithm such that the actions of the players converge to NE by communicating to each other through a connected network. As agents may deviate from their optimal strategies dictated by the NE seeking protocol, we investigate robustness of the proposed algorithm against time‐varying disturbances. In particular, we provide rigorous robustness guarantees by proving input‐to‐state stability (ISS) and ‐stability properties of the NE seeking dynamics. A major concern in communicative schemes among strategic agents is that their private information may be revealed to other agents or to a curious third party who can eavesdrop the communications. Motivated by this, we investigate privacy properties of the algorithm and identify to what extent privacy is preserved when all communicated variables are compromised. Finally, we demonstrate practical applications of our theoretical findings on two case studies; namely, on an energy consumption game and a coordinated charging of electric vehicles.

MSPM: A modularized and scalable multi-agent reinforcement learning-based system for financial portfolio management.

Financial portfolio management (PM) is one of the most applicable problems in reinforcement learning (RL) owing to its sequential decision-making nature. However, existing RL-based approaches rarely focus on scalability or reusability to adapt to the ever-changing markets. These approaches are rigid and unscalable to accommodate the varying number of assets of portfolios and increasing need for heterogeneous data input. Also, RL agents in the existing systems are ad-hoc trained and hardly reusable for different portfolios. To confront the above problems, a modular design is desired for the systems to be compatible with reusable asset-dedicated agents. In this paper, we propose a multi-agent RL-based system for PM (MSPM). MSPM involves two types of asynchronously-updated modules: Evolving Agent Module (EAM) and Strategic Agent Module (SAM). An EAM is an information-generating module with a Deep Q-network (DQN) agent, and it receives heterogeneous data and generates signal-comprised information for a particular asset. An SAM is a decision-making module with a Proximal Policy Optimization (PPO) agent for portfolio optimization, and it connects to multiple EAMs to reallocate the corresponding assets in a financial portfolio. Once been trained, EAMs can be connected to any SAM at will, like assembling LEGO blocks. With its modularized architecture, the multi-step condensation of volatile market information, and the reusable design of EAM, MSPM simultaneously addresses the two challenges in RL-based PM: scalability and reusability. Experiments on 8-year U.S. stock market data prove the effectiveness of MSPM in profit accumulation by its outperformance over five different baselines in terms of accumulated rate of return (ARR), daily rate of return (DRR), and Sortino ratio (SR). MSPM improves ARR by at least 186.5% compared to constant rebalanced portfolio (CRP), a widely-used PM strategy. To validate the indispensability of EAM, we back-test and compare MSPMs on four different portfolios. EAM-enabled MSPMs improve ARR by at least 1341.8% compared to EAM-disabled MSPMs.

Honest agents in a corrupt equilibrium

AbstractWe present a principal–auditor–agent model with adverse selection where the agent can bribe the auditor and the auditor can extort the agent. We show that it is optimal to deter extortion but allow bribery, and we analyze the impact of honest agents in this setting. Unlike strategic agents, honest agents refuse to bribe and therefore cannot reap the benefit of bribery. This unwillingness to bribe raises the cost of participation of honest agents and creates a negative externality by offering a scope for rent to all strategic agents. This negative externality is larger when an agent's bargaining strength is higher. We find that the principal's payoff is nonmonotonic in honesty, and increasing honesty is helpful only if it is widespread enough, leading to a potential corruption trap. We also show that an increase in productivity lowers the threshold for honesty to be beneficial.

How to guarantee food supply during pandemics? Rethinking local food systems from peri-urban strategic agents’ behaviour: The case study of the Barcelona Metropolitan Region

The Covid-19 pandemic has acted as a warning for the world’s current food system, especially in urban contexts with global food dependence. This article aims to analyse the food system behaviour of the Barcelona Metropolitan Region (in the northeast of Spain) during the first stage of the pandemic by deepening the behaviour of different peri-urban agricultural areas in which local food supply is promoted. Semi-structured interviews to 11 entities and institutions located in the peri-urban area of the BMR based on its productive and management profile have been carried out from March to May 2020. The results obtained highlight the socio-economic, environmental, and health perspective of food supply during the pandemic. Main results show 1) shortcomings in the operation and logistics of the metropolitan food system; 2) the complicity between the local producer and the urban consumer through new sales and distribution initiatives, 3) the role of peri-urban agricultural areas for ensuring food supply and land preservation, and 4) the need to initiate cooperation and mutual aid activities between the different agents involved in the food system. Furthermore, agents underlined the need for rethinking the agroeconomic model to strengthening the producer-consumer nexus and promoting local food policy based on food sustainability, sovereignty, and governance.

Counterfactuals with Latent Information

We describe a methodology for making counterfactual predictions in settings where the information held by strategic agents and the distribution of payoff-relevant states of the world are unknown. The analyst observes behavior assumed to be rationalized by a Bayesian model, in which agents maximize expected utility, given partial and differential information about the state. A counterfactual prediction is desired about behavior in another strategic setting, under the hypothesis that the distribution of the state and agents’ information about the state are held fixed. When the data and the desired counterfactual prediction pertain to environments with finitely many states, players, and actions, the counterfactual prediction is described by finitely many linear inequalities, even though the latent parameter, the information structure, is infinite dimensional. (JEL D44, D82, D83)

(Bad) reputation in relational contracting

Motivated by markets for “expertise,” we study a bandit model where a principal chooses between a safe and risky arm. A strategic agent controls the risky arm and privately knows whether its type is high or low. Irrespective of type, the agent wants to maximize duration of experimentation with the risky arm. However, only the high type arm can generate value for the principal. Our main insight is that reputational incentives can be exceedingly strong unless both players coordinate on maximally inefficient strategies on path. We discuss implications for online content markets, term limits for politicians, and experts in organizations.