Sequential Pattern Mining for Situation and Behavior Prediction in Simulated Robotic Soccer

Abstract

Abstract. Agents in dynamic environments have to deal with world representations that change over time. In order to allow agents to act autonomously and to make their decisions on a solid basis an interpretation of the current scene is necessary. If intentions of other agents or events that are likely to happen in the future can be recognized the agent's performance can be improved as it can adapt the behavior to the situation. In this work we present an approach which applies unsupervised symbolic learning off-line to a qualitative abstraction in order to create frequent patterns in dynamic scenes. These patterns can be later applied during runtime in order to predict future situations and behaviors. The pattern mining approach was applied to two games of the 2D RoboCup simulation league.

A new sequential classification to assist Ad auction agent in making decisions

“…Based on our experimental results, we agree with [4] [5] [6] [8] [7] [10] [22] and believe that equipping agents with data mining techniques can lead to more intelligent agents. …”

“…Therefore recent researches combine agent technology and data mining. For example, Lattner and et al [4] mapped quantitative data to qualitative representation. They used an unsupervised off-line learning algorithm to create a set of prediction rules in dynamic environments.…”

Tracking behaviours of cooperative robots within multi-agent domains

“…The discovery of tactical or team behaviors needs the tracking of both the positions of players at any instant of the game and relevant relations able to represent particular interactions between players. Nevertheless, the tracking task becomes very complex because the dynamic conditions of the game brings about drastic changes of positions and interactions between players, which difficult the construction of models capable of recognizing and discovering behaviors of teams playing soccer matches (Lattner et al, 2005). In (Ramos & Ayanegui, 2008b), we proposed a model able to manage the constant changes occurring in the game, which consists in building topological structures based on triangular planar graphs.…”

Mining Temporal Patterns to Improve Agents Behavior: Two Case Studies

“…But to our knowledge, few works have been published on using sequential pattern mining for agent learning. For example, [14] proposed to implement sequential pattern mining in a robot playing soccer. In this case, sequential patterns are used to derive prediction rules about what actions or situations might occur if some preconditions are satisfied.…”