Execution Order Matters in Greedy Algorithms with Limited Information

In this work, we study the multi-agent decision problem where agents try to coordinate to optimize a given system-level objective. While solving for the global optimal is intractable in many cases, the greedy algorithm is a wellstudied and efficient way to provide good approximate solutions notably for submodular optimization problems. Executing the greedy algorithm requires the agents to be ordered and execute a local optimization based on the solutions of the previous agents. However, in limited information settings, passing the solution from the previous agents may be nontrivial, as some agents may not be able to directly communicate with each other. Thus the communication time required to execute the greedy algorithm is closely tied to the order that the agents are given. In this work, we characterize interplay between the communication complexity and agent orderings by showing that the complexity using the best ordering is O(n) and increases considerably to O(n) when using the worst ordering. Motivated by this, we also propose an algorithm that can find an ordering and execute the greedy algorithm quickly, in a distributed fashion. We also show that such an execution of the greedy algorithm is advantageous over current methods for distributed submodular maximization.

[1]  John K. Ousterhout,et al.  In Search of an Understandable Consensus Algorithm , 2014, USENIX ATC.

[2]  Andreas Krause,et al.  Distributed Submodular Maximization , 2014, J. Mach. Learn. Res..

[3]  Andreas Krause,et al.  Efficient Planning of Informative Paths for Multiple Robots , 2006, IJCAI.

[4]  Laurent Bulteau,et al.  Disorders and Permutations , 2021, CPM.

[5]  Andreas Krause,et al.  Near-Optimal Sensor Placements in Gaussian Processes: Theory, Efficient Algorithms and Empirical Studies , 2008, J. Mach. Learn. Res..

[6]  Kyungbaek Kim,et al.  A Survey about Consensus Algorithms Used in Blockchain , 2018, J. Inf. Process. Syst..

[7]  J. Nash.  Equilibrium Points in N-Person Games. , 1950, Proceedings of the National Academy of Sciences of the United States of America.

[8]  Soummya Kar,et al.  Distributed Consensus Algorithms in Sensor Networks: Quantized Data and Random Link Failures , 2007, IEEE Transactions on Signal Processing.

[9]  L. Shapley,et al.  Potential Games , 1994 .

[10]  Yuval Filmus,et al.  The Power of Local Search: Maximum Coverage over a Matroid , 2012, STACS.

[11]  P. Erdos,et al.  On the evolution of random graphs , 1984 .

[12]  Radha Poovendran,et al.  A submodular optimization framework for leader selection in linear multi-agent systems , 2011, IEEE Conference on Decision and Control and European Control Conference.

[13]  Nathan Michael,et al.  Distributed matroid-constrained submodular maximization for multi-robot exploration: theory and practice , 2018, Auton. Robots.

[14]  Sanjeev Sharma,et al.  Robot Path Planning using Swarm Intelligence: A Survey , 2013 .

[15]  Jason R. Marden,et al.  Autonomous Vehicle-Target Assignment: A Game-Theoretical Formulation , 2007 .

[16]  Jan Vondrák,et al.  Maximizing a Monotone Submodular Function Subject to a Matroid Constraint , 2011, SIAM J. Comput..

[17]  Andreas Krause,et al.  Streaming submodular maximization: massive data summarization on the fly , 2014, KDD.

[18]  Hui Lin,et al.  A Class of Submodular Functions for Document Summarization , 2011, ACL.

[19]  Bahman Gharesifard,et al.  On distributed submodular maximization with limited information , 2016, 2016 American Control Conference (ACC).

[20]  Takeo Kanade,et al.  Distributed cosegmentation via submodular optimization on anisotropic diffusion , 2011, 2011 International Conference on Computer Vision.

[21]  M. L. Fisher,et al.  An analysis of approximations for maximizing submodular set functions—I , 1978, Math. Program..

[22]  Jure Leskovec,et al.  Inferring networks of diffusion and influence , 2010, KDD.

[23]  Khaled Ghédira,et al.  Multi Agent model based on Chemical Reaction Optimization with Greedy algorithm for Flexible Job shop Scheduling Problem , 2017, KES.

[24]  Andreas Krause,et al.  Cost-effective outbreak detection in networks , 2007, KDD '07.

[25]  J. Wardrop.  ROAD PAPER. SOME THEORETICAL ASPECTS OF ROAD TRAFFIC RESEARCH. , 1952 .

[26]  Heinrich von Stackelberg.  Market Structure and Equilibrium , 2010 .

[27]  João Pedro Hespanha,et al.  Impact of information in greedy submodular maximization , 2017, 2017 IEEE 56th Annual Conference on Decision and Control (CDC).

[28]  Eitan Zemel,et al.  Nash and correlated equilibria: Some complexity considerations , 1989 .

[29]  Kecheng Liu,et al.  A multi-agent decision support system for stock trading , 2002, IEEE Netw..

[30]  Susel Fernández,et al.  Optimized Sensor Network and Multi-Agent Decision Support for Smart Traffic Light Management , 2018, Sensors.

[31]  Kaveh Madani,et al.  Game theory and water resources , 2010 .

[32]  Bo Zhao,et al.  A Survey on Application of Swarm Intelligence Computation to Electric Power System , 2006, 2006 6th World Congress on Intelligent Control and Automation.

[33]  Dengfeng Sun,et al.  Jacobi-Style Iteration for Distributed Submodular Maximization , 2020, ArXiv.

[34]  Ariel Dinar,et al.  Cooperative water management and ecosystem protection under scarcity and drought in arid and semiarid regions. , 2016 .

[35]  Solmaz S. Kia,et al.  Distributed Strategy Selection: A Submodular Set Function Maximization Approach , 2021, ArXiv.

[36]  George J. Pappas,et al.  Optimal Algorithms for Submodular Maximization with Distributed Constraints , 2019, ArXiv.