• ETRI Journal
• /
• v.43 no.6
• /
• pp.1004-1012
• /
• 2021

The collaboration productively interacting between multi-agents has become an emerging issue in real-world applications. In reinforcement learning, multi-agent environments present challenges beyond tractable issues in single-agent settings. This collaborative environment has the following highly complex attributes: sparse rewards for task completion, limited communications between each other, and only partial observations. In particular, adjustments in an agent's action policy result in a nonstationary environment from the other agent's perspective, which causes high variance in the learned policies and prevents the direct use of reinforcement learning approaches. Unexpected social loafing caused by high dispersion makes it difficult for all agents to succeed in collaborative tasks. Therefore, we address a paradox caused by the social loafing to significantly reduce total returns after a certain timestep of multi-agent reinforcement learning. We further demonstrate that the collaborative paradox in multi-agent environments can be avoided by our proposed effective early stop method leveraging a metric for social loafing.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.945-953
• /
• 2006

The researchers in the field of cognitive science and learning science suggest that the teaching activity induces the elaborative and meaningful learning. Actually, lots of research findings have shown the beneficial effect of learning by teaching such as peer tutoring. But peer tutoring has some limitations in the practical learning context. To overcome some limitations, the new concept of "learning by teaching" through the agent called Teachable Agent. The teachable agent is a modified version of traditional intelligent tutoring system that assigns a role of tutor to teach the agent. The teachable agent monitors individual difference and provides a student with a chance for deep learning and motivation to learn by allowing them to play an active role in the process of learning. That is, The teaching activity induces the elaborative and meaningful learning. This study compared the effects of our teachable agent, KORI, and peer tutoring on the cognition and motivation. The field experiment was conducted to examine whether learning by teaching the teachable agent would be more effective than peer tutoring and reading condition. In the experiment, all participants took 30 minutes lesson on rock and rock cycle together to acquire the base knowledge in the domain. After the lesson, participants were randomly assigned to one of the three experimental conditions; reading condition, peer tutoring condition, and teachable agent condition. Next, participants of each condition moved into separated place and performed their own learning activity. After finishing all of the learning activities in each condition, all participants were instructed to rate the interestingness using a 5-point scale on their own learning activity and leaning material, and were given the comprehension test. The results indicated that the teachable agent condition and the peer tutoring condition showed more interests in the learning than the reading condition. It is suggested that teachable agent has more advantages in overcoming the several practical limitations of peer tutoring such as restrictions in time and place, tutor's cognitive burden, unnecessary interaction during peer tutoring. The applicability and prospects of the teachable agent as an efficient substitute for peer tutoring and traditional intelligent tutoring system were also discussed.

• Journal of Information Processing Systems
• /
• v.16 no.5
• /
• pp.1048-1063
• /
• 2020

In this study, we propose a reinforcement learning agent to control the data transmission rates of nodes in carrier sensing multiple access with collision avoidance (CSMA/CA)-based wireless networks. We design a reinforcement learning (RL) agent, based on Q-learning. The agent learns the environment using the timeout events of packets, which are locally available in data sending nodes. The agent selects actions to control the data transmission rates of nodes that adjust the modulation and coding scheme (MCS) levels of the data packets to utilize the available bandwidth in dynamically changing channel conditions effectively. We use the ns3-gym framework to simulate RL and investigate the effects of the parameters of Q-learning on the performance of the RL agent. The simulation results indicate that the proposed RL agent adequately adjusts the MCS levels according to the changes in the network, and achieves a high throughput comparable to those of the existing data transmission rate adaptation schemes such as Minstrel.

• ETRI Journal
• /
• v.28 no.4
• /
• pp.461-474
• /
• 2006

This project aims to design and develop a prototype for an agent that support online collaborative learning. Online collaborative learning, which has emerged as a new form of education in the knowledge-based society, is regarded as an effective method for improving practical and highly advanced problem-solving abilities. Collaborative learning involves complicated processes, such as organizing teams, setting common goals, performing tasks, and evaluating the outcome of team activities. Thus, a teacher may have difficulty promoting and evaluating the entire process of collaborative learning, and a system may need to be developed to support it. Therefore, to promote interaction among learners in the process of collaborative learning, this study designed an extensible collaborative learning agent (ECOLA) for an online learning environment.

• 제어로봇시스템학회:학술대회논문집
• /
• 1999.10a
• /
• pp.292-297
• /
• 1999

Evolving in artificial agent is an extremely difficult problem, but on the other hand, a challenging task. At present the studies mainly centered on single agent learning problem. In our case, we use simulated soccer to investigate multi-agent cooperative learning. Consider the fundamental differences in learning mechanism, existing reinforcement learning algorithms can be roughly classified into two types-that based on evaluation functions and that of searching policy space directly. Genetic Programming developed from Genetic Algorithms is one of the most well known approaches belonging to the latter. In this paper, we give detailed algorithm description as well as data construction that are necessary for learning single agent strategies at first. In following step moreover, we will extend developed methods into multiple robot domains. game. We investigate and contrast two different methods-simple team learning and sub-group loaming and conclude the paper with some experimental results.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.149.5-149
• /
• 2001

The robot soccer simulation game is a dynamic multi-agent environment. In this paper we suggest a new reinforcement learning approach to each agent´s dynamic positioning in such dynamic environment. Reinforcement learning is the machine learning in which an agent learns from indirect, delayed reward an optimal policy to choose sequences of actions that produce the greatest cumulative reward. Therefore the reinforcement learning is different from supervised learning in the sense that there is no presentation of input-output pairs as training examples. Furthermore, model-free reinforcement learning algorithms like Q-learning do not require defining or learning any models of the surrounding environment. Nevertheless ...

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.173.5-173
• /
• 2001

Q-Learning is a most widely used reinforcement learning, which addresses the question of how an autonomous agent can learn to choose optimal actions to achieve its goal about any one problem. Q-Learning can acquire optimal control strategies from delayed rewards, even when the agent has no prior knowledge of the effects of its action in the environment. If agent has an ability using previous knowledge, then it is expected that the agent can speed up learning by interacting with environment. We present a novel reinforcement learning method using domain knowledge, which is represented by problem-independent features and their classifiers. Here neural network are implied as knowledge classifiers. To show that an agent using domain knowledge can have better performance than the agent with standard Q-Learner. Computer simulations are ...

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.36 no.4
• /
• pp.130-137
• /
• 2013

In this paper, we consider a two-agent scheduling with sequence-dependent exponential learning effects consideration, where two agents A and B have to share a single machine for processing their jobs. The objective function for agent A is to minimize the total completion time of jobs for agent A subject to a given upper bound on the objective function of agent B, representing the makespan of jobs for agent B. By assuming that the learning ratios for all jobs are the same, we suggest an enumeration-based backward allocation scheduling for finding an optimal solution and exemplify it by using a small numerical example. This problem has various applications in production systems as well as in operations management.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.2
• /
• pp.205-211
• /
• 2024

The purpose of this study is to implement an agent that intelligently performs tracking and movement through reinforcement learning using the Unity and ML-Agents. In this study, we conducted an experiment to compare the learning performance between training one agent in a single learning simulation environment and parallel training of several agents simultaneously in a multi-learning simulation environment. From the experimental results, we could be confirmed that the parallel training method is about 4.9 times faster than the single training method in terms of learning speed, and more stable and effective learning occurs in terms of learning stability.

• Proceedings of the IEEK Conference
• /
• 2009.05a
• /
• pp.94-96
• /
• 2009

In this paper, we present the Q-learning method for adaptive traffic signal control on the basis of multi-agent technology. The structure is composed of sixphase agents and one intersection agent. Wireless communication network provides the possibility of the cooperation of agents. As one kind of reinforcement learning, Q-learning is adopted as the algorithm of the control mechanism, which can acquire optical control strategies from delayed reward; furthermore, we adopt dynamic learning method instead of static method, which is more practical. Simulation result indicates that it is more effective than traditional signal system.