Analyzing and Learning an Opponent’s Strategies in the RoboCup Small Size League

Yasui, Kotaro; Kobayashi, Kunikazu; Murakami, Kazuhito; Naruse, Tadashi

doi:10.1007/978-3-662-44468-9_15

Cited by 10 publications

(10 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, they are able to generalize and classify unknown opponents as combination of known ones. Yasui et al [15] introduce a "dissimilarity function" to categorize opponent strategies via cluster analysis. The authors improve their team performance by analyzing logged data of previous matches and showing that team attacking strategies can be recognized and correctly classified.…”

Section: Policy Adaptation In Robocupmentioning

confidence: 99%

Using Monte Carlo Search with Data Aggregation to Improve Robot Soccer Policies

Riccio

Capobianco

Nardi

2017

RoboCup 2016: Robot World Cup XX

View full text Add to dashboard Cite

RoboCup soccer competitions are considered among the most challenging multi-robot adversarial environments, due to their high dynamism and the partial observability of the environment. In this paper we introduce a method based on a combination of Monte Carlo search and data aggregation (MCSDA) to adapt discrete-action soccer policies for a defender robot to the strategy of the opponent team. By exploiting a simple representation of the domain, a supervised learning algorithm is trained over an initial collection of data consisting of several simulations of human expert policies. Monte Carlo policy rollouts are then generated and aggregated to previous data to improve the learned policy over multiple epochs and games. The proposed approach has been extensively tested both on a soccer-dedicated simulator and on real robots. Using this method, our learning robot soccer team achieves an improvement in ball interceptions, as well as a reduction in the number of opponents' goals. Together with a better performance, an overall more efficient positioning of the whole team within the field is achieved.

show abstract

Section: Policy Adaptation In Robocupmentioning

confidence: 99%

Using Monte Carlo Search with Data Aggregation to Improve Robot Soccer Policies

Riccio

Capobianco

Nardi

2017

RoboCup 2016: Robot World Cup XX

View full text Add to dashboard Cite

show abstract

“…Finally online opponent strategy is predicted via clustering. K.Yausui et al [16] propose a dissimilarity function that classifies opponent strategies based upon cluster analysis.…”

Section: Related Workmentioning

confidence: 99%

Opponent Classification in Robot Soccer

Larik

Haider

2015

Current Approaches in Applied Artificial Intelligence

View full text Add to dashboard Cite

Abstract. The paper presents an approach to perform post-hoc analysis of RoboCup Soccer Simulation 3D teams via log files of their matches and to learn a model to classify them not only as being strong, medium or weak but also through their game playing styles such as frequent kickers, frequent dribblers, heavy/lean attackers, etc. The learned model can then be used to further cluster teams to predict game style of similar opponents. We have applied the presented approach to 22 teams from RoboCup 2011 in a fully automated fashion and the results show the validity of our approach. The synergetic effect of these individual actions would emerge as a team strategy and would reveal the quality of the team. This paper presents a novel framework that analyzes key agent behaviors in RoboCup Soccer Simulation 3D[6] using a set of pre-defined rules and classifies a team based upon these extracted team quality features. Team quality features were identified by a group of experts and extraction of these features is automated via a parser application. These features include team players positioning, frequency of players in attacking region, ball possession, average velocity of team players, frequency of kicks executed by teams, shots attempted towards goal etc. Once a team is classified upon its strength parameters then grouping of teams that play similar is done via clustering that would aid in counter strategy formulation against similar opponents. The major

show abstract

“…This method, however, is not suitable for a first match because both the SVM and NN are kinds of supervised machine learning algorithms that require a good training data set for learning. Erdogan et al [4] and Yasui et al [5,6] proposed unsupervised learning algorithms based on clustering. They both use agglomerative hierarchical clustering, but they focus on different data sets: trajectories [4] and deployments [5,6], respectively.…”

mentioning

confidence: 99%

“…Erdogan et al [4] and Yasui et al [5,6] proposed unsupervised learning algorithms based on clustering. They both use agglomerative hierarchical clustering, but they focus on different data sets: trajectories [4] and deployments [5,6], respectively. However, using such geometric data sets can result in data explosion.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Online Strategy Clustering Based on Action Sequences in RoboCupSoccer Small Size League

Adachi¹,

Ito²,

Naruse³

2019

Robotics

Self Cite

View full text Add to dashboard Cite

This paper addresses a strategy learning problem in the RoboCupSoccer Small Size League (SSL). We propose a novel method based on action sequences to cluster an opponent’s strategies online. Our proposed method is composed of the following three steps: (1) extracting typical actions from geometric data to make action sequences, (2) calculating the dissimilarity of the sequences, and (3) clustering the sequences by using the dissimilarity. This method can reduce the amount of data used in the clustering process; handling action sequences instead of geometric data as data-set makes it easier to search actions. As a result, the proposed clustering method is online feasible and also is applicable to countering an opponent’s strategy. The effectiveness of the proposed method was validated by experimental results.

show abstract

Analyzing and Learning an Opponent’s Strategies in the RoboCup Small Size League

Cited by 10 publications

References 4 publications

Using Monte Carlo Search with Data Aggregation to Improve Robot Soccer Policies

Using Monte Carlo Search with Data Aggregation to Improve Robot Soccer Policies

Opponent Classification in Robot Soccer

Online Strategy Clustering Based on Action Sequences in RoboCupSoccer Small Size League

Contact Info

Product

Resources

About