Planning spatial networks with Monte Carlo tree search

被引:5
|
作者
Darvariu, Victor-Alexandru [1 ,2 ]
Hailes, Stephen [1 ]
Musolesi, Mirco [1 ,2 ,3 ]
机构
[1] UCL, London, England
[2] Alan Turing Inst, London, England
[3] Univ Bologna, Bologna, Italy
来源
PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCIENCES | 2023年 / 479卷 / 2269期
基金
英国工程与自然科学研究理事会;
关键词
planning; spatial networks; Monte Carlo tree search; ROBUSTNESS; GO;
D O I
10.1098/rspa.2022.0383
中图分类号
O [数理科学和化学]; P [天文学、地球科学]; Q [生物科学]; N [自然科学总论];
学科分类号
07 ; 0710 ; 09 ;
摘要
We tackle the problem of goal-directed graph construction: given a starting graph, finding a set of edges whose addition maximally improves a global objective function. This problem emerges in many transportation and infrastructure networks that are of critical importance to society. We identify two significant shortcomings of present reinforcement learning methods: their exclusive focus on topology to the detriment of spatial characteristics (which are known to influence the growth and density of links), as well as the rapid growth in the action spaces and costs of model training. Our formulation as a deterministic Markov decision process allows us to adopt the Monte Carlo tree search framework, an artificial intelligence decision-time planning method. We propose improvements over the standard upper confidence bounds for trees (UCT) algorithm for this family of problems that addresses their single-agent nature, the trade-off between the cost of edges and their contribution to the objective, and an action space linear in the number of nodes. Our approach yields substantial improvements over UCT for increasing the efficiency and attack resilience of synthetic networks and real-world Internet backbone and metro systems, while using a wall clock time budget similar to other search-based algorithms. We also demonstrate that our approach scales to significantly larger networks than previous reinforcement learning methods, since it does not require training a model.
引用
收藏
页数:21
相关论文
共 50 条
  • [21] Multi-objective synthesis planning by means of Monte Carlo Tree search
    Lai, Helen
    Kannas, Christos
    Hassen, Alan Kai
    Granqvist, Emma
    Westerlund, Annie M.
    Clevert, Djork-Arne
    Preuss, Mike
    Genheden, Samuel
    ARTIFICIAL INTELLIGENCE IN THE LIFE SCIENCES, 2025, 7
  • [22] A TUTORIAL INTRODUCTION TO MONTE CARLO TREE SEARCH
    Fu, Michael C.
    2020 WINTER SIMULATION CONFERENCE (WSC), 2020, : 1178 - 1193
  • [23] Approximation Methods for Monte Carlo Tree Search
    Aksenov, Kirill
    Panov, Aleksandr, I
    PROCEEDINGS OF THE FOURTH INTERNATIONAL SCIENTIFIC CONFERENCE INTELLIGENT INFORMATION TECHNOLOGIES FOR INDUSTRY (IITI'19), 2020, 1156 : 68 - 74
  • [24] Interpretability of rectangle packing solutions with Monte Carlo tree search
    Lopez, Yeray Galan
    Garcia, Cristian Gonzalez
    Diaz, Vicente Garcia
    Valdez, Edward Rolando Nunez
    Gomez, Alberto Gomez
    JOURNAL OF HEURISTICS, 2024, 30 (3-4) : 173 - 198
  • [25] Fittest survival: an enhancement mechanism for Monte Carlo tree search
    Zhang, Jiajia
    Sun, Xiaozhen
    Zhang, Dandan
    Wang, Xuan
    Qi, Shuhan
    Qian, Tao
    INTERNATIONAL JOURNAL OF BIO-INSPIRED COMPUTATION, 2021, 18 (02) : 122 - 130
  • [26] On Monte Carlo Tree Search and Reinforcement Learning
    Vodopivec, Tom
    Samothrakis, Spyridon
    Ster, Branko
    JOURNAL OF ARTIFICIAL INTELLIGENCE RESEARCH, 2017, 60 : 881 - 936
  • [27] Multiple Pass Monte Carlo Tree Search
    McGuinness, Cameron
    2016 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 2016, : 1555 - 1561
  • [28] Combining Monte-Carlo Tree Search with Proof-Number Search
    Doe, Elliot
    Winands, Mark H. M.
    Soemers, Dennis J. N. J.
    Browne, Cameron
    2022 IEEE CONFERENCE ON GAMES, COG, 2022, : 206 - 212
  • [29] Monte Carlo Tree Search as an intelligent search tool in structural design problems
    Rossi, Leonardo
    Winands, Mark H. M.
    Butenweg, Christoph
    ENGINEERING WITH COMPUTERS, 2022, 38 (04) : 3219 - 3236
  • [30] Monte Carlo Tree Search: a review of recent modifications and applications
    Swiechowski, Maciej
    Godlewski, Konrad
    Sawicki, Bartosz
    Mandziuk, Jacek
    ARTIFICIAL INTELLIGENCE REVIEW, 2023, 56 (03) : 2497 - 2562