Gliders2d: Source Code Base for RoboCup 2D Soccer Simulation League

Prokopenko, Mikhail; Wang, Peter

doi:10.1007/978-3-030-35699-6_33

Cited by 13 publications

(22 citation statements)

References 31 publications

(32 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This observation may appear to contradict other examples where the end-to-end learning has been found to offer advantages over machine learning solutions using a human-defined structure (Lecun et al, 2006 ; Collobert et al, 2011 ; Mnih et al, 2013 ; Bojarski et al, 2016 ). However, the specific problem of learning defensive behaviors considered in this paper was affected by a strong prior optimization within a well-defined structure of Gliders2d , a baseline agent code used by the world champion teams Gliders and Fractals (Prokopenko and Wang, 2019a , b ).…”

Section: Discussionmentioning

confidence: 99%

“…On one hand, many approaches utilize human-selected features and expert-designed strategies, including Situation Based Strategic Positioning (Reis et al, 2001 ), multi-agent positioning mechanism (Akiyama and Noda, 2008 ), coordination system based on setplays (Mota and Reis, 2007 ), positioning based on Delaunay Triangulation (Akiyama and Noda, 2007 ), and Voronoi diagrams (Prokopenko and Wang, 2017 ). Others involve well-optimized defense and attack behaviors in popular code bases such as Agent2d (Akiyama and Nakashima, 2013 ) and Gliders2d (Prokopenko and Wang, 2019a , b ). On the other hand, machine learning approaches have been applied in RCSS environment as well, e.g., a reinforcement learning approach (Riedmiller et al, 2001 , 2008 ; Gabel et al, 2009 ), online planning with tree search method (Akiyama et al, 2012 ), and MAXQ value function decomposition for online planning (Bai et al, 2015 ).…”

Section: Background and Frameworkmentioning

confidence: 99%

See 1 more Smart Citation

Structure-Preserving Imitation Learning With Delayed Reward: An Evaluation Within the RoboCup Soccer 2D Simulation Environment

Nguyen

Prokopenko

2020

Front. Robot. AI

Self Cite

View full text Add to dashboard Cite

We describe and evaluate a neural network-based architecture aimed to imitate and improve the performance of a fully autonomous soccer team in RoboCup Soccer 2D Simulation environment. The approach utilizes deep Q-network architecture for action determination and a deep neural network for parameter learning. The proposed solution is shown to be feasible for replacing a selected behavioral module in a well-established RoboCup base team, Gliders2d, in which behavioral modules have been evolved with human experts in the loop. Furthermore, we introduce an additional performance-correlated signal (a delayed reward signal), enabling a search for local maxima during a training phase. The extension is compared against a known benchmark. Finally, we investigate the extent to which preserving the structure of expert-designed behaviors affects the performance of a neural network-based solution.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Background and Frameworkmentioning

confidence: 99%

Structure-Preserving Imitation Learning With Delayed Reward: An Evaluation Within the RoboCup Soccer 2D Simulation Environment

Nguyen

Prokopenko

2020

Front. Robot. AI

Self Cite

View full text Add to dashboard Cite

show abstract

“…Team Fractals2019 is based on recently released Gliders2d code base [40]. The second version of Gliders2d is described and traced in this study against a pool of benchmark opponents, using a fitness function weighted by relative strengths of the benchmarks.…”

Section: Discussionmentioning

confidence: 99%

“…Each solution is typically evaluated against a specific opponent, over thousands of games, with the fitness function being the average goal difference, while the average points and standard error provide tie-breakers [40]. In other words, a design point (possibly conditioned on the name of a specific opponent) is accepted only if it outperforms every single opponent in the pool of available opponents.…”

Section: Gliders2d: Version V2mentioning

confidence: 99%

See 1 more Smart Citation

Fractals2019: Combinatorial Optimisation with Dynamic Constraint Annealing

Prokopenko¹

2019

Lecture Notes in Computer Science

Self Cite

View full text Add to dashboard Cite

Fractals2019 started as a new experimental entry in the RoboCup Soccer 2D Simulation League, based on Gliders2d code base, and advanced to become a RoboCup-2019 champion. We employ combinatorial optimisation methods, within the framework of Guided Self-Organisation, with the search guided by local constraints. We present examples of several tactical tasks based on the Gliders2d code (version v2), including the search for an optimal assignment of heterogeneous player types, as well as blocking behaviours, offside trap, and attacking formations. We propose a new method, Dynamic Constraint Annealing, for solving dynamic constraint satisfaction problems, and apply it to optimise thermodynamic potential of collective behaviours, under dynamically induced constraints.

show abstract

Engineering Features to Improve Pass Prediction in Soccer Simulation 2D Games

Zare

Sarvmaili

Aref

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Soccer Simulation 2D (SS2D) is a simulation of a real soccer game in two dimensions. In soccer, passing behavior is an essential action for keeping the ball in possession of our team and creating goal opportunities. Similarly, for SS2D, predicting the passing behaviors of both opponents and our teammates helps manage resources and score more goals. Therefore, in this research, we have tried to address the modeling of passing behavior of soccer 2D players using Deep Neural Networks (DNN) and Random Forest (RF). We propose an embedded data extraction module that can record the decision-making of agents in an online format. Afterward, we apply four data sorting techniques for training data preparation. After, we evaluate the trained models' performance playing against 6 top teams of RoboCup 2019 that have distinctive playing strategies. Finally, we examine the importance of different feature groups on the prediction of a passing strategy. All results in each step of this work prove our suggested methodology's effectiveness and improve the performance of the pass prediction in Soccer Simulation 2D games ranging from 5% (e.g., playing against the same team) to 10% (e.g., playing against Robocup top teams).

show abstract

Gliders2d: Source Code Base for RoboCup 2D Soccer Simulation League

Cited by 13 publications

References 31 publications

Structure-Preserving Imitation Learning With Delayed Reward: An Evaluation Within the RoboCup Soccer 2D Simulation Environment

Structure-Preserving Imitation Learning With Delayed Reward: An Evaluation Within the RoboCup Soccer 2D Simulation Environment

Fractals2019: Combinatorial Optimisation with Dynamic Constraint Annealing

Engineering Features to Improve Pass Prediction in Soccer Simulation 2D Games

Contact Info

Product

Resources

About