• The Journal of Korea Robotics Society
• /
• v.14 no.2
• /
• pp.81-86
• /
• 2019

Robots are used in various industrial sites, but traditional methods of operating a robot are limited at some kind of tasks. In order for a robot to accomplish a task, it is needed to find and solve accurate formula between a robot and environment and that is complicated work. Accordingly, reinforcement learning of robots is actively studied to overcome this difficulties. This study describes the process and results of learning and solving which applied reinforcement learning. The mission that the robot is going to learn is bottle flipping. Bottle flipping is an activity that involves throwing a plastic bottle in an attempt to land it upright on its bottom. Complexity of movement of liquid in the bottle when it thrown in the air, makes this task difficult to solve in traditional ways. Reinforcement learning process makes it easier. After 3-DOF robotic arm being instructed how to throwing the bottle, the robot find the better motion that make successful with the task. Two reward functions are designed and compared the result of learning. Finite difference method is used to obtain policy gradient. This paper focuses on the process of designing an efficient reward function to improve bottle flipping motion.

• The Journal of Information Systems
• /
• v.30 no.2
• /
• pp.1-27
• /
• 2021

Purpose Even if many startups firms have developed innovative items and a potential for success, they often have a limited financial resources, which makes them difficult to do business. To overcome this financial difficulty, startups have used one of fintech services, called crowdfunding that can be a good alternative to solving the difficulty of financing. The purpose of this study is to empirically validate the proposed research model that investigates the reasons of trusting crowdfunding platform, which positively leads to two outcomes - intention to participate and word-of-mouth for reward-based crowdfunding project. Design/methodology/approach We proposed several factors categorized as trust, information quality, and platform traits that have a positive impact on trust of crowdfunding platform, which positively leads to intention to participate and word-of-mouth of crowdfunding. The collected(n=285) from individuals who have participated in crowdfunding project was analyzed with SmartPLS 3.0 to test proposed hypotheses. Findings The results showed that all proposed variables (website reputation, crowdfunding familiarity, digital storytelling, information quality, and interaction) had a significant impact on crowfunding platform trust with exception of product differentiation. In addition, crowfunding platform trust was positively associated with participating intention and word-of-mouth. Based on findings, we discussed the research results and implication alone with a direction for future studies.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.22 no.5
• /
• pp.63-68
• /
• 2022

Autonomous driving technology has recently made great progress with the introduction of deep reinforcement learning. In order to effectively use deep reinforcement learning, it is important to select the appropriate activation function. In the meantime, many activation functions have been presented, but they show different performance depending on the environment to be applied. This paper compares and evaluates the performance of 12 activation functions to see which activation functions are effective when using reinforcement learning to learn autonomous driving on highways. To this end, a performance evaluation method was presented and the average reward value of each activation function was compared. As a result, when using GELU, the highest average reward could be obtained, and SiLU showed the lowest performance. The average reward difference between the two activation functions was 20%.

• East Asian Journal of Business Economics (EAJBE)
• /
• v.11 no.1
• /
• pp.1-19
• /
• 2023

Purpose - In this study, we aimed to theorize blockchain-based social media platform users' commitment by examining the impact of their perceived justice of the token reward system. In addition, this study applied psychological ownership theory to verify the underlying mechanism between users' perceptions of justice and their commitment to the platforms. Research design, data, and methodology - To empirically test our conceptual framework in the study, we collected data through a web-based survey approach from the responses of 385 users who had experience with blockchain-based social media platforms. We employed a structural equation modeling approach to empirically test our proposed hypotheses. Result - The results indicated that distributive justice and informational justice have positive effects on user commitment. The results also showed that psychological ownership plays an important role in mediating the relationship between users' sense of distributive justice and commitment, and between procedural justice and commitment. The findings provided a better understanding of the sense of justice and user commitment in a blockchain-based social media environment. Conclusion - This study represents a preliminary attempt to theorize and empirically examine blockchain-based social media platform users' commitment. This study provided important contributions to the literature on how the effect of users' sense of justice in a reward system affects their commitment to blockchain-based social media platforms.

• Journal of information and communication convergence engineering
• /
• v.20 no.3
• /
• pp.181-188
• /
• 2022

An intelligent distributed multi-agent system (IDMS) using reinforcement learning (RL) is a challenging and intricate problem in which single or multiple agent(s) aim to achieve their specific goals (sub-goal and final goal), where they move their states in a complex and cluttered environment. The environment provided by the IDMS provides a cumulative optimal reward for each action based on the policy of the learning process. Most actions involve interacting with a given IDMS environment; therefore, it can provide the following elements: a starting agent state, multiple obstacles, agent goals, and a cluttered index. The reward in the environment is also reflected by RL-based agents, in which agents can move randomly or intelligently to reach their respective goals, to improve the agent learning performance. We extend different cases of intelligent multi-agent systems from our previous works: (a) a proposed environment-clutter-based-index for agent sub-goal selection and analysis of its effect, and (b) a newly proposed RL reward scheme based on the environmental clutter-index to identify and analyze the prerequisites and conditions for improving the overall system.

• CRM연구
• /
• v.2 no.1
• /
• pp.53-72
• /
• 2009

Customer's reward program is a prevailing promotional technique. Recently, both management and marketing fields have been interested in the failure of customer's reward program. However, there are few empirical research regarding this. Therefore, this study examined a research model that employs justice in processing of customer's reward program perceived by customer to explain commitment(calculative commitment, affective commitment) and complaining behavior which is mediated by negative affect. Data was collected from the customers who dissatisfied with their reward programs. For the analysis, frequency, cronbach' ${\alpha}$ and path analysis were used as statistical test tool. Additionally, SPSS 12.0 and AMOS 4.0 were used for analyzing the hypotheses. As a result, proposed structural model largely supports the hypothesized framework and the major findings of this study are summarized as follows: First, distributive and interactional justice were negatively related to negative affect. But procedural justice didn't influence negative affect. Second, negative affect was negatively related to calculative commitment. But affective commitment wasn't influenced by negative affect. Third, negative affect was positively related to complaining behavior. Fourth, calculative commitment was negatively related to complaining behavior. But negative affect didn't influence complaining behavior. In conclusion, It can be posited that justice, negative affect, 2 forms of commitment and complaining behavior are important factors.

• Culinary science and hospitality research
• /
• v.19 no.3
• /
• pp.291-304
• /
• 2013

This study analyzed the effects of Employee Value Proposition (EVP) and the impotance of a job rotation system on the subjective career success. The total 379 samples were surveyed from employees engaging in domestic hotel enterprises located in Seoul using convenient sampling method. The result of this research is as followings. First, EVP has total 5 factors('career development,' 'affiliation,' 'work environment,' 'work content' and 'pay & reward') and job rotation has total 3 factors('individual capacity improvement,' 'procedural justice,' and 'career development'). Second, the results of hypotheses test using a series of multiple regression analysis indicate that EVP factors including 'career development,' 'affiliation,' 'work environment,' and 'work content' influence subjective career success. However, EVP factor of 'pay & reward' does not influence subjective career success. Similarly, EVP factors excluding 'pay & reward' affect a job rotation system. Lastly, a job rotation system positively affects subjective career success. Based on the analysis results, we could draw the importance of the non-financial reward instead of financial reward in the perception of employees' subjective career success. As a research implication, the importance of the creative organization culture was suggested in the conclusion section.

• The Journal of Korean Academy of Sensory Integration
• /
• v.20 no.3
• /
• pp.38-47
• /
• 2022

Objective : We investigated how sensory processing patterns contribute to temperament and character traits in undergraduate students. Methods : A total of 107 undergraduate students were recruited in September 2022 via convenient sampling method. They completed the Adolescent/Adult Sensory Profile and the Temperament and Character Inventory. Multiple regression models were applied to analyze the effect of sensory processing quadrants (low registration, sensation seeking, sensory sensitivity, sensation avoiding) on each temperament (novelty seeking, harm avoidance, reward dependence, persistence) and character (self-directedness, cooperativeness, self-transcendence) traits. Results : Sensation seeking significantly predicted high levels of novelty seeking, reward dependence, persistence, self-directedness, and self-transcendence but low harm avoidance. Low registration predicted high harm avoidance but low levels of reward dependence, persistence, and self-directedness. Reward dependence was predicted by high sensory sensitivity and low sensation avoiding. Conclusion : This study demonstrated that sensory processing patterns affected novelty seeking, harm avoidance, reward dependence, persistence, self-directedness, and self-transcendence in young adults.

• Journal of The Korean Association For Science Education
• /
• v.26 no.2
• /
• pp.222-231
• /
• 2006

The purposes of this study were to survey internal engagement in science fair and explore the causal relationship between internal engagement and motivation for participation. A written questionnaire on queries into motivation for participation and internal engagement were developed and tested. The subjects were 1066 students from 4th to 9th grade who had participated in the Youth Science Contest under the auspices of the Korea Science Foundation. Interest and commitment were selected as constructing factors of internal engagement. Through exploratory factor analysis, preference, reward, and social motivation were determined to be the factors affecting the motivation to participate. Boys showed higher internal engagement than girls, and interest and commitment were found to be higher in elementary school students(p<0.01). There was no difference in interest among elementary school students; however, fourth grade students showed lower commitment than other students(p<0.01). Ninth grade students showed the lowest interest and commitment among junior high school students(p<0.01). To explore the causal relationship between internal engagement and factors influence internal engagement, path analysis was used. The selected model illustrated how reward motivation affected commitment directly, and how preference motivation affected interest directly but only commitment indirectly through interest. Reward motivation affected commitment with a standardized direct effect coefficient of 0.17. Preference motivation affected interest with a standardized direct effect coefficient of 0.75 and commitment with a standardized total effect coefficient of 0.63(direct effect; 0.27 and indirect effect; 0.36). In addition, interest affected commitment with a standardized direct effect coefficient of 0.49. Social motivation did not affect interest and commitment and reward motivation did not affect interest.

• International Journal of Control, Automation, and Systems
• /
• v.1 no.3
• /
• pp.358-367
• /
• 2003

We consider discrete-time factorial Markov Decision Processes (MDPs) in multiple decision-makers environment for infinite horizon average reward criterion with a general joint reward structure but a factorial joint state transition structure. We introduce the "localization" concept that a global MDP is localized for each agent such that each agent needs to consider a local MDP defined only with its own state and action spaces. Based on that, we present a gradient-ascent like iterative distributed algorithm that converges to a local optimal solution of the global MDP. The solution is an autonomous joint policy in that each agent's decision is based on only its local state.cal state.