Results 1 -
6 of
6
Using focal point learning to improve tactic coordination in human-machine interactions
- In IJCAI 2007
, 2007
"... We consider an automated agent that needs to coordinate with a human partner when communication between them is not possible or is undesirable (tactic coordination games). Specifically, we examine situations where an agent and human attempt to coordinate their choices among several alternatives with ..."
Abstract
-
Cited by 4 (1 self)
- Add to MetaCart
We consider an automated agent that needs to coordinate with a human partner when communication between them is not possible or is undesirable (tactic coordination games). Specifically, we examine situations where an agent and human attempt to coordinate their choices among several alternatives with equivalent utilities. We use machine learning algorithms to help the agent predict human choices in these tactic coordination domains. Learning to classify general human choices, however, is very difficult. Nevertheless, humans are often able to coordinate with one another in communication-free games, by using focal points, “prominent ” solutions to coordination problems. We integrate focal points into the machine learning process, by transforming raw domain data into a new hypothesis space. This results in classifiers with an improved classification rate and shorter training time. Integration of focal points into learning algorithms also results in agents that are more robust to changes in the environment. 1
Using focal points learning to improve human-machine tactic coordination
- JAAMAS
"... Abstract We consider an automated agent that needs to coordinate with a human partner when communication between them is not possible or is undesirable (tacit coordination games). Specifically, we examine situations where an agent and human attempt to coordinate their choices among several alternat ..."
Abstract
-
Cited by 3 (3 self)
- Add to MetaCart
(Show Context)
Abstract We consider an automated agent that needs to coordinate with a human partner when communication between them is not possible or is undesirable (tacit coordination games). Specifically, we examine situations where an agent and human attempt to coordinate their choices among several alternatives with equivalent utilities. We use machine learning algorithms to help the agent predict human choices in these tacit coordination domains. Experiments have shown that humans are often able to coordinate with one another in communication-free games, by using focal points, "prominent" solutions to coordination problems. We integrate focal point rules into the machine learning process, by transforming raw domain data into a new hypothesis space. We present extensive empirical results from three different tacit coordination domains. The Focal Point Learning approach results in classifiers with a 40% to 80% higher correct classification rate, and shorter training time, than when using regular classifiers, and a 35% higher correct classification rate than classical focal point techniques without learning. In addition, the integration of focal points into learning algorithms results in agents that are more robust to changes in the environment. We also present several results describing various biases that might arise in Focal Point based coordination.
The Limited Utility of Communication in Simple Organisms
"... Many forms of communication have evolved in the animal kingdom for different purposes. In this paper we investigate the limits of communication for simple reactive organisms and show that communication has only limited benefits in bi-ologically inspired foraging tasks and can even have detri-mental ..."
Abstract
-
Cited by 1 (0 self)
- Add to MetaCart
(Show Context)
Many forms of communication have evolved in the animal kingdom for different purposes. In this paper we investigate the limits of communication for simple reactive organisms and show that communication has only limited benefits in bi-ologically inspired foraging tasks and can even have detri-mental effects in certain environments. Based on these re-sults, we argue that simple agents with simple architectures need very special environmental conditions for communica-tion to benefit them and thus to evolve.
Probabilistic Roadmaps for Unknown Kinodynamic Constraints
, 2006
"... Probabilistic Roadmap (PRM) planners have been used to generate paths for articulated robots for several years. By using random sampling techniques, PRM based planners are able to plot paths for robots with many degrees of freedom without needing to explore large parts of the search space that tradi ..."
Abstract
- Add to MetaCart
(Show Context)
Probabilistic Roadmap (PRM) planners have been used to generate paths for articulated robots for several years. By using random sampling techniques, PRM based planners are able to plot paths for robots with many degrees of freedom without needing to explore large parts of the search space that traditional planners would have to examine to create efficient paths. This has enabled them to be used with robots operating in high dimensional spaces, which are common in multi-agent robotics. Some limitations exist with PRM planners when they need to work with robots that are constrained in their motion or when several robots are involved in the plan. Current work has enabled planners with prior knowledge of agent constraints to predict how motion limitations will affect the robot in different poses. By adapting and extending PRM based algorithms to remove the need to know about constraints beforehand, the research proposed here aims to improve the capabilities of PRM methods and enable them to be used in more domains than is currently possible. Various algorithms using PRM are discussed in detail and new ideas on how some can be extended are outlined as well as describing the work that has already gone towards implementing such a system.
THE COST OF COMMUNICATION: EFFICIENT COORDINATION IN MULTI-AGENT TERRITORY EXPLORATION TASKS
, 2006
"... by ..."
(Show Context)
