A Monge Sequence-Based Approach to Characterize the Competitive Newsvendor Problem

Replicating cash flows of multiple agents in game-theoretic settings tends to be a challenging task. In this paper, we consider the competitive newsvendor game where multiple newsvendors choose inventory levels before demand arrival and the unmet demand of each newsvendor spills over to multiple other newsvendors. We show that this spillover behavior and the resulting cash flows of each newsvendor can be replicated within a transportation problem after assigning artificial costs on spillover behavior. This replication provides an opportunity to study structural properties of the problem, as well as determine the equilibrium of the game. This paradigm of using artificial costs within an optimization framework to replicate agents’ cash flows can be used in many other games as well.

Agen Ganda dan Perubahan Praktik Pertanian

This article aims to explain the role of dual agents in bringing about changes in agricultural practices. Starting from the phenomenon of changing plant species that occurred in Parbotihan Village, Onan Ganjang District, Humbang Hasundutan Regency. Changes in these types of crops often occur in a short period of time and are followed by most farmers. These changes sometimes occur without considering the adequacy of land, cultivation knowledge, and capital capacity. So, often the changes that occur are not profitable for them. On the other hand, these changes also change the cultivation pattern which requires farmers to learn from the beginning as a consequence of changing the types of plants. This research was conducted during the research period of my thesis by collecting data through observation and interviews. The selected informants are farmers who are involved in changing practices. The author found that the role of multiple agents such as relatives, friends, skippers, and group leaders with different capacities had a role in influencing farmers' actions. Multiple agents act as initiators, motivators, introductors, educators, and interventors. This research also shows that the social arena as public space becomes an effective arena in exchanging information and influences that encourage practice change.

Stacking Subsidies in Factor Markets: Evidence from Market Experiments

Government policies employ different support programs such as subsidies to reduce risks, increase efficiency in markets, and enhance societal welfare. In markets such as ethanol markets, where multiple agents receive subsidy, it is often difficult to determine whether recipients of these support programs will transfer some of their payments to other agents in the market. In this study, we use laboratory market experiments to understand subsidy incidence in markets where both buyers and sellers receive subsidies, and there are few buyers relative to sellers. Our results show that when subsidizing both sides of the market, framing effects matter, and when markets are buyer concentrated, subsidy distributions generally tend to favor buyers. With a per-unit subsidy of 20 tokens to both sides and an equal number of buyers and sellers in the market, we find that buyers increase their earnings by 13.4% while seller earnings decrease by 16.1%. On a per-schedule basis, buyer earnings in the concentrated market are similar to what we observed in the competitive market.

Multi-itinerary optimization as cloud service

In this paper, we describe multi-itinerary optimization (MIO)---a novel Bing Maps service that automates the process of building itineraries for multiple agents while optimizing their routes to minimize travel time or distance. MIO can be used by organizations with a fleet of vehicles and drivers, mobile salesforce, or a team of personnel in the field, to maximize workforce efficiency. It supports a variety of constraints, such as service time windows, duration, priority, pickup and delivery dependencies, and vehicle capacity. MIO also considers traffic conditions between locations, resulting in algorithmic challenges at multiple levels (e.g., calculating time-dependent travel-time distance matrices at scale and scheduling services for multiple agents). To support an end-to-end cloud service with turnaround times of a few seconds, our algorithm design targets a sweet spot between accuracy and performance. Toward that end, we build a scalable approach based on the ALNS metaheuristic. Our experiments show that accounting for traffic significantly improves solution quality: MIO finds efficient routes that avoid late arrivals, whereas traffic-agnostic approaches result in a 15% increase in the combined travel time and the lateness of an arrival. Furthermore, our approach generates itineraries with substantially higher quality than a cutting-edge heuristic (LKH), with faster running times for large instances.

Decentralized Multi-Agent Control of a Manipulator in Continuous Task Learning

Many real-world tasks require multiple agents to work together. When talking about multiple agents in robotics, it is usually referenced to multiple manipulators in collaboration to solve a given task, where each one is controlled by a single agent. However, due to the increasing development of modular and re-configurable robots, it is also important to investigate the possibility of implementing multi-agent controllers that learn how to manage the manipulator’s degrees of freedom (DoF) in separated clusters for the execution of a given application (e.g., being able to face faults or, partially, new kinematics configurations). Within this context, this paper focuses on the decentralization of the robot control action learning and (re)execution considering a generic multi-DoF manipulator. Indeed, the proposed framework employs a multi-agent paradigm and investigates how such a framework impacts the control action learning process. Multiple variations of the multi-agent framework have been proposed and tested in this research, comparing the achieved performance w.r.t. a centralized (i.e., single-agent) control action learning framework, previously proposed by some of the authors. As a case study, a manipulation task (i.e., grasping and lifting) of an unknown object (to the robot controller) has been considered for validation, employing a Franka EMIKA panda robot. The MuJoCo environment has been employed to implement and test the proposed multi-agent framework. The achieved results show that the proposed decentralized approach is capable of accelerating the learning process at the beginning with respect to the single-agent framework while also reducing the computational effort. In fact, when decentralizing the controller, it is shown that the number of variables involved in the action space can be efficiently separated into several groups and several agents. This simplifies the original complex problem into multiple ones, efficiently improving the task learning process.

A2CM: a new multi-agent algorithm

<p class="Abstract">Reinforcement learning is currently one of the most researched fields of artificial intelligence. New algorithms are being developed that use neural networks to compute the selected action, especially for deep reinforcement learning. One subcategory of reinforcement learning is multi-agent reinforcement learning, in which multiple agents are present in the world. As it involves the simulation of an environment, it can be applied to robotics as well. In our paper, we use our modified version of the advantage actor–critic (A2C) algorithm, which is suitable for multi-agent scenarios. We test this modified algorithm on our testbed, a cooperative–competitive pursuit–evasion environment, and later we address the problem of collision avoidance.</p>

Study about Law Multi-Issue Automatic Negotiation Method Based on Artificial Intelligence and Multiagent Evolutionary Algorithm

Multiagent System (MAS) is a self-learning intelligent system formed by many single agents. Each agent in the MAS works independently of the other, and they have all the characteristics of an agent system. It can respond to changes and countermeasures based on its own external environmental conditions. When solving a complex problem, multiple agents can form a group to solve the problem together. In this paper, the agent’s evolutionary algorithm is integrated into the actual problem—multi-issue autonegotiation of law. According to this problem, the autonegotiation solution process and corresponding model are designed. In addition, a new type of solution is proposed for multiple legal issues. Compared with traditional solutions, the applicability has great advantages. Among them, the autonegotiation result of all the agent’s total utility can be quickly found. In the changing environment, this article focuses on the multiagent system negotiation problem. According to the distributed information sharing of multiple agents, even if the case reveals incomplete information, the multiagent can be generated while ignoring the incomplete information. Optimal solution is proposed. The experimental results show that the success rate of the system in analyzing multiple legal issues and autonegotiations reached 67.56% under the condition of incomplete information from the outside world.