• Journal of Information Processing Systems
• /
• 제16권5호
• /
• pp.1223-1230
• /
• 2020

Recently, extensive studies have been conducted to apply deep learning to reinforcement learning to solve the state-space problem. If the state-space problem was solved, reinforcement learning would become applicable in various fields. For example, users can utilize dance-tutorial systems to learn how to dance by watching and imitating a virtual instructor. The instructor can perform the optimal dance to the music, to which reinforcement learning is applied. In this study, we propose a method of reinforcement learning in which the action space is dynamically adjusted. Because actions that are not performed or are unlikely to be optimal are not learned, and the state space is not allocated, the learning time can be shortened, and the state space can be reduced. In an experiment, the proposed method shows results similar to those of traditional Q-learning even when the state space of the proposed method is reduced to approximately 0.33% of that of Q-learning. Consequently, the proposed method reduces the cost and time required for learning. Traditional Q-learning requires 6 million state spaces for learning 100,000 times. In contrast, the proposed method requires only 20,000 state spaces. A higher winning rate can be achieved in a shorter period of time by retrieving 20,000 state spaces instead of 6 million.

• 대한의용생체공학회:의공학회지
• /
• 제17권3호
• /
• pp.395-401
• /
• 1996

In this paper a new neural network model, time delayed counterpropagation neural networks (TDCPN) which have high recognition rate and short total learning time, is proposed for electromyogram(EMG) recognition. Signals the proposed model increases the recognition rates after learned the regional temporal correlation of patterns using time delay properties in input layer, and decreases the learning time by using winner-takes-all learning rule. The ouotar learning rule is put at the output layer so that the input pattern is able to map a desired output. We test the performance of this model with EMG signals collected from a normal subject. Experimental results show that the recognition rates of the suggested model is better and the learning time is shorter than those of TDNN and CPN.

• 4차산업연구
• /
• 제4권2호
• /
• pp.1-9
• /
• 2024

Purpose: While most studies focus on learning styles in face-to-face education, research on online learning environments, especially by age in lifelong education, is limited. This study aims to propose a direction for online learning by analyzing digital literacy and e-Learning learning styles by age in lifelong education. Research design, data and methodology: The study surveyed 100 online learners from an open university in Seoul. Using an e-Learning learning styles test, frequency analysis was conducted by gender, age, and digital literacy. A learning plan was then proposed based on the results. Results: The study found no age-related differences in digital literacy. Both men and women shared similar ratios of Environment-dependent and self-directed learning styles, reflecting the characteristics of online learners using digital devices. Conclusions: In lifelong education, e-Learning design should accommodate diverse learning styles: web/app designs for Environment-independent and self-directed learners, short/long formats for Passive learners, real-time (LMS)/non-real-time (ZOOM) systems for Positive and cooperative learners, and AI/human tutors for Environment-dependent and self-directed learners.

• 대한치위생과학회지
• /
• 제1권2호
• /
• pp.21-39
• /
• 2018

The purpose of this study was to examine the learning competencies of students at a college from September 1 to November 30, 2017, in an effort to provide some information on how to foster learning competencies in college years, which lay the foundation for work and social lives. 1. The learning competencies of the subjects consisted of academic vision, student identity, cognitive regulation, emotional regulation, learning management and creating learning environments. Out of five points, they scored the highest in academic vision and student identity with 3.34, followed by learning management with 3.20, creating learning environments with 3.18, emotional regulation with 3.16 and cognitive regulation with 3.14. 2. There were statistically significant differences in academic vision according to age, the area of major, the academic credential of their fathers, commuting time, military service experience and career plans. 3. There were statistically significant differences in student identity and cognitive regulation according to gender, age, the area of major, the academic credential of their fathers, commuting time, military service experience and career plans. 4. There were statistically significant differences in emotional regulation according to age, the area of major, the academic credential of their fathers, commuting time, career plans and daily mean study hours. 5. There were statistically significant differences in learning management according to gender, age, the area of major, grade point average, the academic credential of their fathers, career plans and daily mean study hours. 6. There were statistically significant differences in creating learning environments according to gender, age, the area of major, the academic credential of fathers, commuting time, career plans and daily mean study hours. As they were poorest at the cognitive regulation area among the areas of learning competencies, self-directed learning programs that deal with how to study, learning process, how to take notes and arrange them, how to link different pieces of acquired knowledge and how to map out study plans should be developed to give support to students.

• International Journal of Internet, Broadcasting and Communication
• /
• 제15권1호
• /
• pp.195-202
• /
• 2023

This study is a comparative survey study conducted to explore the differences in learners' learning flow and learning satisfaction according to the non-face-to-face class operation methods implemented at universities. After implementing different class management methods for the same subject taught by the same instructor non-face-to-face for 15 weeks, each learning flow and learning satisfaction were compared and analyzed, and the collected data were analyzed with IBM SPSS 21.0. As a result of the study, learning flow was high in the order of lectures using real-time ZOOM and recorded lectures using self-studio(3.41±0.91, 3.28±1.01), and learning satisfaction was high in the order of lectures using real-time ZOOM and lectures using the automatic recording system of classes(3.40±0.80, 3.30±0.74). The item with the lowest score was the PPT audio recording lecture in both areas of learning flow and learning satisfaction(2.72±1.04, 1.73±1.04). Considering that system errors such as sound in the smart lecture environment operated for the first time in this study affected the research results, it is suggested that future research should be conducted by supplementing the corresponding part.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제8권1호
• /
• pp.31-36
• /
• 2008

This paper presents adaptive learning data of evolvable neural networks (ENNs) for time series prediction of nonlinear dynamic systems. ENNs are a special class of neural networks that adopt the concept of biological evolution as a mechanism of adaptation or learning. ENNs can adapt to an environment as well as changes in the enviromuent. ENNs used in this paper are L-system and DNA coding based ENNs. The ENNs adopt the evolution of simultaneous network architecture and weights using indirect encoding. In general just previous data are used for training the predictor that predicts future data. However the characteristics of data and appropriate size of learning data are usually unknown. Therefore we propose adaptive change of learning data size to predict the future data effectively. In order to verify the effectiveness of our scheme, we apply it to chaotic time series predictions of Mackey-Glass data.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.920-924
• /
• 2005

This paper presents adaptive learning data of evolvable neural networks (ENNs) for time series prediction of nonlinear dynamic systems. ENNs are a special class of neural networks that adopt the concept of biological evolution as a mechanism of adaptation or learning. ENNs can adapt to an environment as well as changes in the environment. ENNs used in this paper are L-system and DNA coding based ENNs. The ENNs adopt the evolution of simultaneous network architecture and weights using indirect encoding. In general just previous data are used for training the predictor that predicts future data. However the characteristics of data and appropriate size of learning data are usually unknown. Therefore we propose adaptive change of learning data size to predict the future data effectively. In order to verify the effectiveness of our scheme, we apply it to chaotic time series predictions of Mackey-Glass data.

• 한국지능시스템학회논문지
• /
• 제14권3호
• /
• pp.363-368
• /
• 2004

In this paper, the robustness property of 2nd-order iterative learning control(ILC) method for a class of linear and nonlinear discrete-time dynamic systems is studied. 2nd-order ILC method has the PD-type learning algorithm based on both time-domain performance and iteration-domain performance. It is proved that the 2nd-order ILC method has robustness in the presence of state disturbances, measurement noise and initial state error. In the absence of state disturbances, measurement noise and initialization error, the convergence of the 2nd-order ILC algorithm is guaranteed. A numerical example is given to show the robustness and convergence property according to the learning parameters.

• International Journal of Internet, Broadcasting and Communication
• /
• 제16권3호
• /
• pp.321-328
• /
• 2024

This paper designs a disease prediction algorithm to diagnose migraine among the types of diseases in advance by learning algorithms using machine learning-based time series analysis. This study utilizes patient data statistics, such as electroencephalogram activity, to design a prediction algorithm to determine the onset signals of migraine symptoms, so that patients can efficiently predict and manage their disease. The results of the study evaluate how accurate the proposed prediction algorithm is in predicting migraine and how quickly it can predict the onset of migraine for disease prevention purposes. In this paper, a machine learning algorithm is used to analyze time series of data indicators used for migraine identification. We designed an algorithm that can efficiently predict and manage patients' diseases by quickly determining the onset signaling symptoms of disease development using existing patient data as input. The experimental results show that the proposed prediction algorithm can accurately predict the occurrence of migraine using machine learning algorithms.

• 대한산업공학회지
• /
• 제32권1호
• /
• pp.61-73
• /
• 2006

This paper deals with traveling salesman problem(TSP) with the stochastic travel time. Practically, the travel time between demand points changes according to day and time zone because of traffic interference and jam. Since the almost pervious studies focus on TSP with the deterministic travel time, it is difficult to apply those results to logistics problem directly. But many logistics problems are strongly related with stochastic situation such as stochastic travel time. We need to develop the efficient solution method for the TSP with stochastic travel time. From the previous researches, we know that Q-learning technique gives us to deal with stochastic environment and neural network also enables us to calculate the Q-value of Q-learning algorithm. In this paper, we suggest an algorithm for TSP with the stochastic travel time integrating Q-learning and neural network. And we evaluate the validity of the algorithm through computational experiments. From the simulation results, we conclude that a new route obtained from the suggested algorithm gives relatively more reliable travel time in the logistics situation with stochastic travel time.