• Journal of Internet of Things and Convergence
• /
• v.7 no.4
• /
• pp.29-34
• /
• 2021

In order to apply reinforcement learning to a robot in a real environment, it is necessary to use simulation in a virtual environment because numerous iterative learning is required. In addition, it is difficult to apply a learning algorithm that requires a lot of computation for a robot with low-spec. hardware. In this study, ML-Agent, a reinforcement learning frame provided by Unity, was used as a virtual simulation environment to apply reinforcement learning to the obstacle collision avoidance problem of mobile robots with low-spec hardware. A DQN supported by ML-Agent is adopted as a reinforcement learning algorithm and the results for a real robot show that the number of collisions occurred less then 2 times per minute.

• The KIPS Transactions:PartB
• /
• v.9B no.2
• /
• pp.139-146
• /
• 2002

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 these algorithms can learn the optimal policy if the agent can visit every state-action pair infinitely. However, the biggest problem of monolithic reinforcement learning is that its straightforward applications do not successfully scale up to more complex environments due to the intractable large space of states. In order to address this problem, we suggest Adaptive Mediation-based Modular Q-Learning (AMMQL) as an improvement of the existing Modular Q-Learning (MQL). While simple modular Q-learning combines the results from each learning module in a fixed way, AMMQL combines them in a more flexible way by assigning different weight to each module according to its contribution to rewards. Therefore in addition to resolving the problem of large state space effectively, AMMQL can show higher adaptability to environmental changes than pure MQL. In this paper we use the AMMQL algorithn as a learning method for dynamic positioning of the robot soccer agent, and implement a robot soccer agent system called Cogitoniks.

• Journal of Korea Multimedia Society
• /
• v.12 no.12
• /
• pp.1819-1826
• /
• 2009

To enhance the automatic performance of existing predicting and planning algorithms that require a predefined probability of the states' transition, this paper proposes a multiple sequence generation system. When interacting with unknown environments, a virtual agent needs to decide which action or action order can result in a good state and determine the transition probability based on the current state and the action taken. We describe a sequential behavior generation method motivated from the change in the agent's state in order to help the virtual agent learn how to adapt to unknown environments. In a sequence learning process, the sensed states are grouped by a set of proposed motivation filters in order to reduce the learning computation of the large state space. In order to accomplish a goal with a high payoff, the learning agent makes a decision based on the observation of states' transitions. The proposed multiple sequence behaviors generation system increases the complexity and heightens the automatic planning of the virtual agent for interacting with the dynamic unknown environment. This model was tested in a virtual library to elucidate the process of the system.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.59-64
• /
• 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 chose 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 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 it can learn the optimal policy if the agent can visit every state- action pair infinitely. However, the biggest problem of monolithic reinforcement learning is that its straightforward applications do not successfully scale up to more complex environments due to the intractable large space of states. In order to address this problem. we suggest Adaptive Mediation-based Modular Q-Learning (AMMQL)as an improvement of the existing Modular Q-Learning (MQL). While simple modular Q-learning combines the results from each learning module in a fixed way, AMMQL combines them in a more flexible way by assigning different weight to each module according to its contribution to rewards. Therefore in addition to resolving the problem of large state effectively, AMMQL can show higher adaptability to environmental changes than pure MQL. This paper introduces the concept of AMMQL and presents details of its application into dynamic positioning of robot soccer agents.

• The Journal of the Korea Contents Association
• /
• v.9 no.10
• /
• pp.51-58
• /
• 2009

Reinforcement learning is a sub area of machine learning concerned with how an agent ought to take actions in an environment so as to maximize some notion of long-term reward. In the case of multi-agent, especially, which state space and action space gets very enormous in compared to single agent, so it needs to take most effective measure available select the action strategy for effective reinforcement learning. This paper proposes a multi-agent reinforcement learning model based on fuzzy inference system in order to improve learning collect speed and select an effective action in multi-agent. This paper verifies an effective action select strategy through evaluation tests based on Robocup Keepaway which is one of useful test-beds for multi-agent. Our proposed model can apply to evaluate efficiency of the various intelligent multi-agents and also can apply to strategy and tactics of robot soccer system.

• Journal of IKEEE
• /
• v.23 no.4
• /
• pp.1187-1194
• /
• 2019

This paper proposes a CNN architecture as value function network of an artificial intelligence Othello game agent and its learning scheme using reinforcement learning algorithm. We propose an approach to construct the value function network by using CNN to learn the records of professional players' real game and an approach to enhance the network parameter by learning from self-play using reinforcement learning algorithm. The performance of value function network CNN was compared with existing ANN by letting two agents using each network to play games each other. As a result, the winning rate of the CNN agent was 69.7% and 72.1% as black and white, respectively. In addition, as a result of applying the reinforcement learning, the performance of the agent was improved by showing 100% and 78% winning rate, respectively, compared with the network-based agent without the reinforcement learning.

• Educational Technology International
• /
• v.19 no.1
• /
• pp.61-91
• /
• 2018

The animated pedagogical agent has been implemented to promote learning outcomes and motivation in multimedia learning. It has been claimed that one of the advantages of using pedagogical agent is persona effect - the personalization or social presence of pedagogical agent can enhance learning engagement and motivation. However, prior research is inconclusive as to whether and how the features of the pedagogical agent have effects on the persona effect. This study investigated whether the similarity between a pedagogical agent and the real instructor in terms of the voice and outlook would improve students' perception of the agent's persona. The study also examined the effect by the size of pedagogical agent on the persona perception. Two experiments were conducted with a total of 115 college students. Experiment 1 indicated a significant main effect of voice on the persona perception. Experiment 2 was conducted to examine whether the size of pedagogical agent would affect the voice effect on the persona perception. The results showed that the instructor-like voice yielded higher persona perception regardless of the pedagogical agent's size. Overall, the study findings indicated that the similarity in voice positively fostered the agent's persona.

• The KIPS Transactions:PartA
• /
• v.12A no.6 s.96
• /
• pp.547-556
• /
• 2005

Online collaborative teaming, which has emerged as a new type of education in knowledge-based society, is being discussed actively in the areas of action learning at companies and project-based learning and inquiry-based learning at schools. It regards as an effective method for improving learners practical and highly advanced problem solving abilities, and for stimulating their absorption into learning through pursuing common goals of learning together. Different from individual learning, however, collaborative learning involves complicated processes such as organizing teams, setting common goals, performing tasks and evaluating the outcome of team activities .Thus, it is difficult for a teacher to promote and evaluate the whole process of collaborative learning, and it is necessary to develop systems to support collaborative learning. Therefore, in order to monitor and promote interaction among learners in the process of collaborative learning, the present study developed an extensible collaborative teaming supporting agent (ECOLA) in online learning environments.

• The Journal of Korean Association of Computer Education
• /
• v.6 no.1
• /
• pp.87-94
• /
• 2003

Recently there are many tries and researches for using web agent in education. The agent is a computer program as a common name which takes a role like a proxy or a middle ware for accomplishing the given something on behalf of user. Lately the agent is being used in the various fields. The teaching and learning is the one of those. It is the web agent that support the teaching and learning on the web. It has a concept of the program or the engine is able to do teachers roles on the behalf of him on the web. Not only the web agent is able to do teachers roles on the behalf of him, but also it is a helper that helps learners on the web. The web based teaching and learning environment has the web agent offers the personal and the adaptive information, interface, or contents.

• Journal of Korea Multimedia Society
• /
• v.19 no.2
• /
• pp.344-351
• /
• 2016

Recently, mobile application for english vocabulary learning is being developed actively. However, most mobile English vocabulary learning applications did not effectively connected with the technical advantages of mobile learning. Also,the study of mobile english vocabulary learning app are still insufficient. Therefore, this paper development a english vocabulary context-learning Agent that can practice context learning more reasonably using a location-based service, a character recognition technology and augmented reality technology based on smart phones. In order to evaluate the performance of the proposed agent, we have measured the precision and usability. As results of experiments, the precision of learning vocabulary is 89% and 'Match between system and the real world', 'User control and freedom', 'Recognition rather than recall', 'Aesthetic and minimalist design' appeared to be respectively 3.91, 3.80, 3.85, 4.01 in evaluation of usability. It were obtained significant results.