We study the fair and efficient allocation of a set of indivisible goods among agents, where each good has several copies, and each agent has an additively separable concave valuation function with a threshold. These valuations capture the property of diminishing marginal returns, and they are more general than the well-studied case of additive valuations. We present a polynomial-time algorithm that approximates the optimal Nash social welfare (NSW) up to a factor of e1/e ≈ 1.445. This matches with the state-of-the-art approximation factor for additive valuations. The computed allocation also satisfies the popular fairness guarantee of envy-freeness up to one good (EF1) up to a factor of 2 + ε. For instances without thresholds, it is also approximately Pareto-optimal. For instances satisfying a large market property, we show an improved approximation factor. Lastly, we show that the upper bounds on the optimal NSW introduced in Cole and Gkatzelis (2018) and Barman et al. (2018) have the same value.

Autonomous systems that operate in the open world often use incomplete models of their environment. Model incompleteness is inevitable due to the practical limitations in precise model specification and data collection about open-world environments. Due to the limited fidelity of the model, agent actions may produce negative side effects (NSEs) when deployed. Negative side effects are undesirable, unmodeled effects of agent actions on the environment. NSEs are inherently challenging to identify at design time and may affect the reliability, usability and safety of the system. We present two complementary approaches to mitigate the NSE via: (1) learning from feedback, and (2) environment shaping. The solution approaches target settings with different assumptions and agent responsibilities. In learning from feedback, the agent learns a penalty function associated with a NSE. We investigate the efficiency of different feedback mechanisms, including human feedback and autonomous exploration. The problem is formulated as a multi-objective Markov decision process such that optimizing the agent’s assigned task is prioritized over mitigating NSE. A slack parameter denotes the maximum allowed deviation from the optimal expected reward for the agent’s task in order to mitigate NSE. In environment shaping, we examine how a human can assist an agent, beyond providing feedback, and utilize their broader scope of knowledge to mitigate the impacts of NSE. We formulate the problem as a human-agent collaboration with decoupled objectives. The agent optimizes its assigned task and may produce NSE during its operation. The human assists the agent by performing modest reconfigurations of the environment so as to mitigate the impacts of NSE, without affecting the agent’s ability to complete its assigned task. We present an algorithm for shaping and analyze its properties. Empirical evaluations demonstrate the trade-offs in the performance of different approaches in mitigating NSE in different settings.

As a complement to conventional AI solutions, emergent intelligence (EI) exhibits competitiveness in 6G IIoT scenario for its various outstanding features including robustness, protection to privacy, and scalability. However, despite the low computational complexity, EI is challenged by its high demand of data traffic in massive deployment. We propose to leverage massive twinning, which 6G is envisaged to support, to reduce the data traffic in EI and therewith enhance its performance.

In the last decade, there have been significant advances in multi-agent reinforcement learning (MARL) but there are still numerous challenges, such as high sample complexity and slow convergence to stable policies, that need to be overcome before wide-spread deployment is possible. However, many real-world environments already, in practice, deploy sub-optimal or heuristic approaches for generating policies. An interesting question that arises is how to best use such approaches as advisors to help improve reinforcement learning in multi-agent domains. In this paper, we provide a principled framework for incorporating action recommendations from online suboptimal advisors in multi-agent settings. We describe the problem of ADvising Multiple Intelligent Reinforcement Agents (ADMIRAL) in nonrestrictive general-sum stochastic game environments and present two novel Q-learning based algorithms: ADMIRAL - Decision Making (ADMIRAL-DM) and ADMIRAL - Advisor Evaluation (ADMIRAL-AE), which allow us to improve learning by appropriately incorporating advice from an advisor (ADMIRAL-DM), and evaluate the effectiveness of an advisor (ADMIRAL-AE). We analyze the algorithms theoretically and provide fixed point guarantees regarding their learning in general-sum stochastic games. Furthermore, extensive experiments illustrate that these algorithms: can be used in a variety of environments, have performances that compare favourably to other related baselines, can scale to large state-action spaces, and are robust to poor advice from advisors.

Estimating value and/or bid distributions from an observed sequence of auctions is a fundamental challenge in Econometrics with direct practical applic ations. For example, these fundamentals allow one to analyze the performance of an auction and make co unterfactual predictions about alternatives. The difficulty of this problem depends on the fo rmat of the auctions and the structure of the observed information from each one, as well as how the fundamentals of bidders are interrelated and vary across the sequence of observations. In this paper, we study a basic version of the afore-describe d estimation challenge, wherein the auction format and the bidder distributions stay fixed across observations, and the bidders have independent private values (which are independently resam pled across different observations). The auction formats that we consider are first-and second-pri ce auctions, as well as Dutch and English auctions. What will make our problem challenging is that (i) our bidders are ex ante asymmetric, drawing their independent private values from different distributions; (ii) we will make no parametric assumptions about these distributions; and (iii) we will only be observing the 1 identity of the winner and the price they paid but not the losi ng bids. Under this observational model and our independent private values assumption above, we can focus our attention on first-and second-price auctions, and our results automatically e xtend to Dutch and English auctions. In the above settings, we give computationally and sample ef ficient methods for estimating all agents' bid distributions and (under equilibrium assumpti ons) value distributions: In the case of first-price auctions, we provide finite-sample es timation guarantees under L evy, Kolmogorov and T otal V ariation distance with minimal assumptions. Under (a condition weaker than) a lower bound on the density of the bid dis tributions (although we actually do not need existence of densities), Theorem 2.2 shows that the bid distributions can be estimated to within ε in L evy distance, using 1/ ε

We all have preferences when multiple choices are available. If we insist on satisfying our preferences only, we may suffer a loss due to conflicts with other people's identical selections. Such a case applies when the choice cannot be divided into multiple pieces due to the intrinsic nature of the resources. Former studies, such as the top trading cycle, examined how to conduct fair joint decision-making while avoiding decision conflicts from the perspective of game theory when multiple players have their own deterministic preference profiles. However, in reality, probabilistic preferences can naturally appear in relation to the stochastic decision-making of humans. Here, we theoretically derive conflict-free joint decision-making that can satisfy the probabilistic preferences of all individual players. More specifically, we mathematically prove the conditions wherein the deviation of the resultant chance of obtaining each choice from the individual preference profile, which we call the loss, becomes zero, meaning that all players' satisfaction is perfectly appreciated while avoiding decision conflicts. Furthermore, even in situations where zero-loss conflict-free joint decision-making is unachievable, we show how to derive joint decision-making that accomplishes the theoretical minimum loss while ensuring conflict-free choices. Numerical demonstrations are also shown with several benchmarks.

A smart city is a city that uses technology to provide services and solve city problems. The main goals of a smart city are to improve policy efficiency, reduce waste and inconvenience, improve social and economic quality, and maximize social inclusion. Due to the breadth of technologies that have been implemented under the smart city label, it is difficult to distill a precise definition of a smart city. As the world's population continues to urbanize – by 2050, 66% of the world's population is expected to be urban – there is a global trend toward the creation of smart cities. This tendency not only causes many physical, social, behavioural, economic, and infrastructure issues, but it also creates many opportunities.

Yellow.ai, which offers automation across customer engagement, support and conversational commerce for enterprises, has announced the availability of pre-built Dynamic AI Agents for rapid deployment across a number of verticals. The agents are designed to connect conversations across voice, text and chat, in multiple languages. The agents, which will be available in Yellow.ai's Agents are also available to enhance employee experience by automating HR processes like onboarding and training, and IT management services. We seem to stand on the brink of a working world in which everything is automated, both for employees and customers.

Interacting particle system (IPS) models have proven to be highly successful for describing the spatial movement of organisms. However, it has proven challenging to infer the interaction rules directly from data. In the field of equation discovery, the Weak form Sparse Identification of Nonlinear Dynamics (WSINDy) methodology has been shown to be very computationally efficient for identifying the governing equations of complex systems, even in the presence of substantial noise. Motivated by the success of IPS models to describe the spatial movement of organisms, we develop WSINDy for second order IPSs to model the movement of communities of cells. Specifically, our approach learns the directional interaction rules that govern the dynamics of a heterogeneous population of migrating cells. Rather than aggregating cellular trajectory data into a single best-fit model, we learn the models for each individual cell. These models can then be efficiently classified according to the active classes of interactions present in the model. From these classifications, aggregated models are constructed hierarchically to simultaneously identify different species of cells present in the population and determine best-fit models for each species. We demonstrate the efficiency and proficiency of the method on several test scenarios, motivated by common cell migration experiments.

RecSimNG is a scalable, modular, differentiable simulator implemented in Edward2 and TensorFlow. It offers: a powerful, general probabilistic programming language for agent-behavior specification; an XLA-based vectorized execution model for running simulations on accelerated hardware; and tools for probabilistic inference and latent-variable model learning, backed by automatic differentiation and tracing. We describe RecSim NG and illustrate how it can be used to create transparent, configurable, end-to-end models of a recommender ecosystem. Specifically, we present a collection of use cases that demonstrate how the functionality described above can help both researchers and practitioners easily develop and train novel algorithms for recommender systems. Please cite the paper if you use the code from this repository in your work. This is not an officially supported Google product.