Distributed Constraint Optimization (DCOP) is an elegant formalism relevant to many areas in multiagent systems, yet complete algorithms have not been pursued for real world applications due perceived complexity. To capably capture a rich class of complex problem domains, we introduce the Multi-Event Scheduling (DiMES) framework and design congruent DCOP formulations with binary constraints which are proven yield optimal solution. approach real-world efficiency requirements, obtain immense...

Bike Sharing Systems (BSSs) are widely adopted in major cities of the world due to concerns associated with extensive private vehicle usage, namely, increased carbon emissions, traffic congestion and usage nonrenewable resources. In a BSS, base stations strategically placed throughout city each station is stocked pre-determined number bikes at beginning day. Customers hire from one return them another station. Due unpredictable movements customers hiring bikes, there either (more than...

Taxis (which include cars working with car aggregation systems such as Uber, Grab, Lyft etc.) have become a critical component in the urban transportation. While most research and applications context of taxis focused on improving performance from customer perspective, this paper, we focus taxi driver perspective. Higher revenues for drivers can help bring more into system thereby availability customers dense cities. Typically, when there is no board, will cruise around to find either...

On-demand ride-pooling (e.g., UberPool, LyftLine, GrabShare) has recently become popular because of its ability to lower costs for passengers while simultaneously increasing revenue drivers and aggregation companies Uber). Unlike in Taxi on Demand (ToD) services – where a vehicle is assigned one passenger at time on-demand ride-pooling, each must serve multiple with heterogeneous origin destination pairs without violating any quality constraints. To ensure near real-time response, existing...

In real-world sequential decision making tasks like autonomous driving, robotics, and healthcare, learning from observed state-action trajectories is critical for imitation, classification, clustering. For example, self-driving cars must replicate human driving behaviors, while robots healthcare systems benefit modeling sequences, whether or not they come expert data. Existing trajectory encoding methods often focus on specific rely reward signals, limiting their ability to generalize across...

Unsupervised Environment Design (UED) formalizes the problem of autocurricula through interactive training between a teacher agent and student agent. The generates new environments with high learning potential, curating an adaptive curriculum that strengthens student's ability to handle unseen scenarios. Existing UED methods mainly rely on regret, metric measures difference agent's optimal actual performance, guide design. Regret-driven generate curricula progressively increase environment...

Distributed POMDPs provide an expressive framework for modeling multiagent collaboration problems, but NEXP-Complete complexity hinders their scalability and application in real-world domains. This paper introduces a subclass of distributed POMDPs, TREMOR, algorithm to solve such POMDPs. The primary novelty TREMOR is that agents plan individually with single agent POMDP solver use social model shaping implicitly coordinate other agents. Experiments demonstrate can solutions orders magnitude...

The use of distributed POMDPs for cooperative teams has been severely limited by the incredibly large joint policyspace that results from combining policy-spaces individual agents. However, much computational cost exploring entire policy space can be avoided observing in many domains important interactions between agents occur a relatively small set scenarios, previously defined as coordination locales (CLs) [11]. Moreover, even when numerous might occur, given policies there are few actual...

This research is motivated by large scale problems in urban transportation and labor mobility where there congestion for resources uncertainty movement. In such domains, even though the individual agents do not have an identity of their own explicitly interact with other agents, they effect agents. While has been much handling implicit effects, it primarily assumed de- terministic movements We address issue decision support that are identical involuntary dynamic environments. For instance, a...

The widespread availability of cell phones has enabled non-profits to deliver critical health information their beneficiaries in a timely manner. This paper describes our work assist that employ automated messaging programs preventive care (new and expecting mothers) during pregnancy after delivery. Unfortunately, key challenge such delivery is significant fraction drop out the program. Yet, often have limited health-worker resources (time) place crucial service calls for live interaction...

This paper describes an innovative multiagent system called SAVES with the goal of conserving energy in commercial buildings. We specifically focus on application to be deployed existing university building that provides several key novelties: (i) jointly performed facility management team, is based actual occupant preferences and schedules, consumption loss data, real sensors hand-held devices, etc.; (ii) it addresses novel scenarios require negotiations groups occupants conserve energy;...

Bike Sharing System (BSS) is a green mode of transportation that employed extensively for short distance travels in major cities the world. Unfortunately, users behaviour driven by their personal needs can often result empty or full base stations, thereby resulting loss customer demand. To counter this demand, BSS operators typically utilize fleet carrier vehicles repositioning bikes between stations. However, fuel burning incurs significant amount routing, labor cost and further increases...

Due to increased traffic congestion and carbon emissions, Bike Sharing Systems (BSSs) are adopted in various cities for short distance travels, specifically last mile transportation. The success of a bike sharing system depends on its ability have bikes available at the "right" base stations times. Typically, carrier vehicles used perform repositioning between so as satisfy customer requests. Owing uncertainty demand day-long repositioning, problem having right times is challenging one. In...

Distributed Partially Observable Markov Decision Problems (Distributed POMDPs) are a popular approach for modeling multi-agent systems acting in uncertain domains. Given the significant complexity of solving distributed POMDPs, particularly as we scale up numbers agents, one has focused on approximate solutions. Though this is efficient, algorithms within do not provide any guarantees solution quality. A second less focuses global optimality, but typical results available only two and also...

Many applications of networks agents, including mobile sensor networks, unmanned air vehicles, autonomous underwater involve 100s agents acting collaboratively under uncertainty. Distributed Partially Observable Markov Decision Problems (Distributed POMDPs) are well-suited to address such applications, but so far, only limited scale-ups up five have been demonstrated. This paper escalates the scale-up, presenting an algorithm called FANS, increasing number in distributed POMDPs for first...

Mass Rapid Transit using rail is a popular mode of transport employed by millions people in many urban cities across the world. Typically, these networks are massive, used and thus, can be soft target for criminals. In this paper, we consider problem scheduling randomised patrols improving security such networks. Similar to existing work protecting critical infrastructure, also employ Stackelberg Games represent problem. solving games massive networks, make two key contributions. Firstly,...

In this paper, we solve cooperative decentralized stochastic planning problems, where the interactions between agents (specified using transition and reward functions) are dependent on number of (and not identity individual agents) involved in interaction. A collision robots a narrow corridor, defender teams coordinating patrol activities to secure target, etc. examples such anonymous interactions. Formally, consider problems that subset well known Decentralized MDP (DEC-MDP) model,...

While Markov Decision Processes (MDPs) have been shown to be effective models for planning under uncertainty, the objective minimize expected cumulative cost is inappropriate high-stake problems. As such, Yu, Lin, and Yan (1998) introduced Risk-Sensitive MDP (RS-MDP) model, where find a policy that maximizes probability within some user-defined threshold. In this paper, we revisit problem introduce new algorithms are based on classical techniques, such as depth-first search dynamic...