• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.493-493
• /
• 2000

Multi-valued decision making for transitional stochastic events was newly derived based on conditional probability of database. The two values (on-off) decision making method without transition had been proposed by one of the author in a previous work for a purpose of realizing human on-off decision making. The current method is an extension of the previous on-off decision making. By combining the conditional probability and the transitional probability, the closed form of the algorithm for the multi-valued transitional decision making was derived. The proposed multi-valued decision making was successfully applied to the determination of the five levels of the vigilance of a subject during the EEG recording; awake stage, drowsy stage and sleeping stages (stage 1, stage 2/3, REM (rapid eye movement)). The method for determining the vigilance level can be directly usable for the two purposes; selection of awake EEG segments for automatic EEG interpretation, and determination of sleep stages through sleep EEG. The proposed multi-valued decision making with a mathematical background of the probability can be applicable widely, in industries and in medical fields for purposes of the multi-valued decision making.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.783-786
• /
• 1997

Automatic realization of on-off human decision making was derived based on a conditional probability. Following the proposed procedure, problems of insulator washing timing in power substations and spike detection on EEG(electroencephalogram) records were appropriately solved.

• Journal of Internet Computing and Services
• /
• v.25 no.4
• /
• pp.7-21
• /
• 2024

There are various machine learning techniques such as Reinforcement Learning, Deep Learning, Neural Network Learning, and so on. In recent, Large Language Models (LLMs) are popularly used for Generative AI based on Reinforcement Learning. It makes decisions with the most optimal rewards through the fine tuning process in a particular situation. Unfortunately, LLMs can not provide any explanation for how they reach the goal because the training is based on learning of black-box AI. Reinforcement Learning as black-box AI is based on graph-evolving structure for deriving enhanced solution through adjustment by human feedback or reinforced data. In this research, for mutually exclusive decision-making, Mutually Exclusive Learning (MEL) is proposed to provide explanations of the chosen goals that are achieved by a decision on both ends with specified conditions. In MEL, decision-making process is based on the tree-based structure that can provide processes of pruning branches that are used as explanations of how to achieve the goals. The goal can be reached by trade-off among mutually exclusive alternatives according to the specific contextual conditions. Therefore, the tree-based structure is adopted to provide feasible solutions with the explanations based on the pruning branches. The sequence of pruning processes can be used to provide the explanations of the inferences and ways to reach the goals, as Explainable AI (XAI). The learning process is based on the pruning branches according to the multi-dimensional contextual conditions. To deep-dive the search, they are composed of time window to determine the temporal perspective, depth of phases for lookahead and decision criteria to prune branches. The goal depends on the policy of the pruning branches, which can be dynamically changed by configured situation with the specific multi-dimensional contextual conditions at a particular moment. The explanation is represented by the chosen episode among the decision alternatives according to configured situations. In this research, MEL adopts the tree-based learning model to provide explanation for the goal derived with specific conditions. Therefore, as an example of mutually exclusive problems, employment process is proposed to demonstrate the decision-making process of how to reach the goal and explanation by the pruning branches. Finally, further study is discussed to verify the effectiveness of MEL with experiments.

• Smart Structures and Systems
• /
• v.20 no.5
• /
• pp.525-537
• /
• 2017

This paper explores different vibration control strategies for the cancellation of human-induced vibration on a structure with time-varying modal parameters. The main motivation of this study is a lively urban stress-ribbon footbridge (Pedro $G\acute{o}mez$ Bosque, Valladolid, Spain) that, after a whole-year monitoring, several natural frequencies within the band of interest (normal paring frequency range) have been tracked. The most perceptible vibration mode of the structure at approximately 1.8 Hz changes up to 20%. In order to find a solution for this real case, this paper takes the annual modal parameter estimates (approx. 14000 estimations) of this mode and designs three control strategies: a) a tuned mass damper (TMD) tuned to the most-repeated modal properties of the aforementioned mode, b) two semi-active TMD strategies, one with an on-off control law for the TMD damping, and other with frequency and damping tuned by updating the damper force. All strategies have been carefully compared considering two structure models: a) only the aforementioned mode and b) all the other tracked modes. The results have been compared considering human-induced vibrations and have helped the authors on making a decision of the most advisable strategy to be practically implemented.

• Proceedings of the KSR Conference
• /
• 2011.10a
• /
• pp.2573-2580
• /
• 2011

SMRT(Seoul Metropolitan Rapid Transit Corporation) changed PC-oriented office system to Smart-Phone oriented mobile office system(STnF: SMRT Talk & Flash), and provided smart phone to all staffs. STnF make it possible not only voice communication between staff, but also to input and inquiry all kind of information related to work(ex: Facility maintenance, Rolling Stock management, office management, Finance, human resource etc.) based on smart phone. STnF conquers the limitation of time and space. All staffs can commute to the site directly to reduce moving time and to share information between staffs, so SMRT improve speed and reliability of work by making a quick decision and action on the spot. After introduction of STnF, within 1 year, the equipment failure ratio fell to 47%. Mobile office system reduced and level-off the step of approval. so enhanced civil customer service. Quantification and scientific movement of work make it fast evolution and development. These business improvement make it possible to put more outstanding manpower on new business and R&D project, and this contributes to company's management effectiveness.

• Korean Journal of Environment and Ecology
• /
• v.29 no.5
• /
• pp.681-692
• /
• 2015

It is well documented that physiological and nutritional condition of wintering birds is strongly related to migration success to breeding sites, and also breeding success. However, how abiotic factors during winter affect the migration and breeding successes still remains unclear. Thus, this study developed a dynamic-state-dependent model for wintering life-history to identify the potential impact on the life-history, success to breeding site and breeding success of wintering birds, which are related to temperature fluctuation, interspecific competition and human disturbance at the wintering sites. To find the best-fit-model, we referred to the existing research data on wintering ecology of Red-crowned cranes (Grus japonensis) in Cheolwon, Korea, which is well documented as a long-term wintering study. Our model predicted that the higher temperature fluctuation and a higher rate of human disturbance are negatively related to migration success to breeding sites and their fitness, ultimately breeding success via changing of proportion in resource allocation (for e. g., lower energy compensation or higher level of stress accumulation). Particularly, the rate of body mass compensation after arrival at wintering sites may be accelerated when there are less temperature fluctuations and a lower rate of human disturbance. In addition, the rate of interspecific competition sharing the wintering foraging sites is negatively related to the rate of body mass compensation. Consequently, we discussed the conservation strategies of wintering birds based on the outcomes of the model.

• Journal of Distribution Science
• /
• v.13 no.4
• /
• pp.45-53
• /
• 2015

Purpose - Online shopping enables consumers to conveniently purchase products irrespective of the time and place. As a result, several online shopping companies have emerged to cater to this growing market and, therefore, the competition among them has become increasingly intense. This paper evaluates the comparative efficiency of online shopping companies for a multi-year period (2009-2013), in order to help online shopping managers identify major drivers for enhancing management efficiency and the subsequent competitiveness. Research design, data, and methodology - The researchers collected the data from 2009 to 2013 from the distribution yearbook. This paper analyzes the marketability (sales figures), profitability (business profits), and management conditions (net profits) of domestic online shopping enterprises by incorporating information on human resources (number of employees) and material resources (total assets and capital). Therefore, the number of employees, total assets, and capital are selected as input variables, and sales figures, business profits, and net profits as the output variables. In this study, Data Envelopment Analysis (DEA) was used to measure the comparative efficiency of domestic online shopping companies. In addition, the Malmquist Productivity Index was used to evaluate the trend of change of Decision Making Units' (DMUs') efficiency for a multi-year period. Results - First, as of 2013, Interpark (2.415) was found to be the most efficient online shopping enterprise, followed by Aladdin Communications (2.117), Hyundai Home shopping (1.867), Home&Shopping (1.176), NS Home shopping (1.170), Commerce Planet (1.126), CJ O Shopping (1.105), Ebay Korea (1.088), and GS Home Shopping (1.051). Second, this study recognizes how the management efficiency has changed for the period 2009-2013. Third, the lesser the capital and employees, the more are the net profits, and the better is the management efficiency of domestic online shopping companies. Lastly, the productivity of such companies is influenced by endogenous factors rather than exogenous factors such as shifts in business environment and technological advances. Conclusions - DHC Korea influenced various distribution channels to reach customers through the Internet. Consequently, this helped in increasing the awareness about its products, in addition to an increase in sales. These achievements can be attributed to the characteristics of online shopping companies. Although it is easy for these companies to suggest goods for one-off purchases, they however have difficulties in retaining customers. Overcoming this challenge can be one of the ways to benchmark a successful case of an efficient company. For example, an online shopping company can attract customers by developing a corresponding mobile application as a convenient way to shop online. Additionally, they can satisfy customers by quick delivery of purchased products, which is possible by building an effective logistics network. Our study indicates that the productivity of an online shopping company was influenced by endogenous factors driven by improvements in managerial practices rather than exogenous factors. Accordingly, online shopping companies should adopt strategies to improve their operational efficiency rather than sales volume-oriented management.