• Environmental Analysis Health and Toxicology
• /
• 제25권4호
• /
• pp.273-278
• /
• 2010

Objectives : This study was aimed to examine the acute toxicity assessment of two new algicides, thiazolidinediones derivatives (TD53 and TD49), which were synthesized to selectively control red tide, to the marine ecosystem. Methods : The assessment employed by a new method using Ulva pertusa Kjellman which has been recently accepted as a standard method of ISO. The toxicity was assessed by calculating the $EC_{50}$ (Effective Concentration of 50%), NOEC (No Observed Effect Concentration) and PNEC (Predicted No Effect Concentration) using acute toxicity data obtained from exposure experiments. $EC_{50}$ value of TD49 and TD53 was examined by 96-hrs exposure together with Solutol as a TD49 dispersing agent and DMSO as a TD53 solvent. Results : $EC_{50}$ value of TD53 was $1.65\;{\mu}M$. From the results, values of NOEC and PNEC were calculated to be $0.63\;{\mu}M$ and 1.65 nM, respectively. DMSO under the range of $0{\sim}10\;{\mu}M$, which is same solvent concentration used in examining TD53, showed no toxic effect. $EC_{50}$ value of TD49 was $0.18\;{\mu}M$ and that of Solutol was $1.70\;{\mu}M$. NOEC and PNEC of TD49 were $0.08\;{\mu}M$ and 0.18 nM, respectively and those for Solutol were $1.25\;{\mu}M$ and 1.25 nM, respectively. Conclusions : From the values of NOEC, PNEC of TD53 and TD49, TD49 showed 9 times stronger toxicity than TD53. On the other hand, DMSO showed no toxicity on the Ulva pertusa Kjellman, but Solutol was found to be a considerable toxicity by itself.

• 정보처리학회논문지B
• /
• 제11B권1호
• /
• pp.77-82
• /
• 2004

강화학습에서 temporal-credit 할당 문제 즉, 에이전트가 현재 상태에서 어떤 행동을 선택하여 상태전이를 하였을 때 에이전트가 선택한 행동에 대해 어떻게 보상(reward)할 것인가는 강화학습에서 중요한 과제라 할 수 있다. 본 논문에서는 조합최적화(hard combinational optimization) 문제를 해결하기 위한 새로운 메타 휴리스틱(meta heuristic) 방법으로, greedy search뿐만 아니라 긍정적 반응의 탐색을 사용한 모집단에 근거한 접근법으로 Traveling Salesman Problem(TSP)를 풀기 위해 제안된 Ant Colony System(ACS) Algorithms에 Q-학습을 적용한 기존의 Ant-Q 학습방범을 살펴보고 이 학습 기법에 다양화 전략을 통한 상태전이와 TD-오류를 적용한 학습방법인 Ant-TD 강화학습 방법을 제안한다. 제안한 강화학습은 기존의 ACS, Ant-Q학습보다 최적해에 더 빠르게 수렴할 수 있음을 실험을 통해 알 수 있었다.

• Journal of Pharmaceutical Investigation
• /
• 제29권4호
• /
• pp.349-353
• /
• 1999

A novel polymeric tablet of tinidazole (TD) was formulated to treat Helicobacter pylori and Giardia lambria more efficiently with reduced hepatotoxicity by controlling the release of TD after oral administration. TD tablets containing various concentrations of either xanthan gum (XG, viscosity enhancer) and/or polycarbophil (PC, mucoadhesive) were prepared by the wet granulation method. In vitro release of TD into pH 2.0 and pH 5.0 buffer solutions was observed at 37°C by using an USP dissolution tester and an UV (313 nm) spectrophotometer. In vivo absorption of TD tablets was investigated in rabbits by measuring the blood concentration of TD after oral administration using a HPLC. Compared to a commercial TD tablet, in vitro release of TD in both pH 2.0 and pH 5.0 buffer solutions significantly decreased as the concentration: of XG or PC in the tablet increased up to 30%. However, when XG and PC was added in combination, TD was completely released in a pH 5.0 buffer solution within 8 hours, whereas the release of TD in pH 2.0 buffer solution significantly decreased. TD in a commercial tablet was rapidly absorbed after oral administration in rabbits. After oral administration of the polymeric tablets that contain both XG and PC, plasma concentration of TD dramatically decreased. Since the oral absorption of TD significantly decreased by the addition of XG and PC in the tablets while TD completely released in a pH 5.0 buffer solution, it was speculated that more TD was retained in the gastrointestinal tract. Thus, it was possible to control the release of TD by changing the content of XG and/or PC in the tablet, thereby manipulating the release rate and the gastrointestinal retention of TD after oral administration in rabbits.

• 인터넷정보학회논문지
• /
• 제17권6호
• /
• pp.25-31
• /
• 2016

최근 무선통신 기술이 급속도로 발전하면서 스마트 센서를 이용한 사물통신 (Machine-to-Machine, M2M)과 사물인터넷 (Internet of Things, IoT) 등의 신규 서비스들이 창출되고 있다. 본 논문에서는 사용자들이 스마트폰이나 태블릿PC 등의 사용자 장비 (User Equipment, UE)를 이용하여 사물인터넷 디바이스 (IoT Device, IoTD)를 제어하는 환경에서, 스몰셀 기지국(Small-cell Access Point, SAP)을 도입하여 IoTD를 개인화하는 방법을 제안한다. 이를 위해, 먼저 UE, IoTD, 그리고 SAP 로 구성되는 시스템 아키텍처를 소개한다. 이후, IoTD의 프로파일 등록 절차와 댁내 외 UE의 IoTD를 제어 절차로 구성되는 IoTD 개인화 방법을 제안한다. 마지막으로 시뮬레이션을 통해 제안하는 방법의 시스템 성능을 분석하였고, 전형적인 방법보다 제안하는 방법이 패킷 지연시간, 패킷 손실률, 그리고 정규화된 수율 측면에서 성능이 향상됨을 알 수 있었다.

• 한국통신학회논문지
• /
• 제30권5B호
• /
• pp.304-309
• /
• 2005

본 논문에서는 TD(temporal differences) 기법을 사용한 localized QoS 라우팅 기법을 제안하였다. 이 기법은 이웃노드로부터 얻어지는 성공 기댓값을 통해 라우팅 정책을 결정하는 기법이다. 이에 본 논문에서는 라우팅 성공 기댓값을 기반으로 한 다양한 탐색기법으로 경로 선택 시 라우팅 성능을 비교 평가하였으며, 특히 Exploration Bonus를 적용한 탐색 기법이 다른 탐색 기법에 비해 더욱 우수한 성능을 보여주고 있는데, 이는 다른 탐색 기법에 비해 네트워크 상황에 더 적응적으로 경로를 선택할 수 있기 때문이다.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.1072-1075
• /
• 2005

This paper presents that TD method is applied to the human adaptive devices for smart home with context awareness (or recognition) technique. For smart home, the very important problem is how the appliances (or devices) can adapt to user. Since there are many humans to manage home appliances (or devices), managing the appliances automatically is difficult. Moreover, making the users be satisfied by the automatically managed devices is much more difficult. In order to do so, we can use several methods, fuzzy controller, neural network, reinforcement learning, etc. Though the some methods could be used, in this case (in dynamic environment), reinforcement learning is appropriate. Among some reinforcement learning methods, we select the Temporal Difference learning method as a core algorithm for adapting the devices to user. Since this paper assumes the environment is a smart home, we simply explained about the context awareness. Also, we treated with the TD method briefly and implement an example by VC++. Thereafter, we dealt with how the devices can be applied to this problem.

• 한국인터넷방송통신학회논문지
• /
• 제10권5호
• /
• pp.265-270
• /
• 2010

강화학습(reinforcement learning)은 동적 환경과 시행-착오를 통해 상호 작용하면서 학습을 수행한다. 그러므로 동적 환경에서 TD-학습과 TD(${\lambda}$)-학습과 같은 강화학습 방법들은 전통적인 통계적 학습 방법보다 더 빠르게 학습을 할 수 있다. 그러나 제안된 대부분의 강화학습 알고리즘들은 학습을 수행하는 에이전트(agent)가 목표 상태에 도달하였을 때만 강화 값(reinforcement value)이 주어지기 때문에 최적 해에 매우 늦게 수렴한다. 본 논문에서는 미로 환경(maze environment)에서 최단 경로를 빠르게 찾을 수 있는 강화학습 방법(GORLS : Goal-Directed Reinforcement Learning System)을 제안하였다. GDRLS 미로 환경에서 최단 경로가 될 수 있는 후보 상태들을 선택한다. 그리고 나서 최단 경로를 탐색하기 위해 후보 상태들을 학습한다. 실험을 통해, GDRLS는 미로 환경에서 TD-학습과 TD(${\lambda}$)-학습보다 더 빠르게 최단 경로를 탐색할 수 있음을 알 수 있다.

• 한국전자파학회논문지
• /
• 제7권4호
• /
• pp.320-327
• /
• 1996

This paper presents that the lossy or amplification property of the Finite Difference-Time Domain(FD-TD) method based on the leap-frog scheme is theoretically verified by using a plane wave analysis. The basic algorithm of the FD-TD method is introduced in order to help understanding the analysis procedure. Since our analysis is formulated by the Von Neumann's approach, the stability inequality is also produced as an another outcome.

• 한국정보처리학회논문지
• /
• 제6권5호
• /
• pp.1303-1311
• /
• 1999

In order to learn in dynamic environments, reinforcement learning algorithms like Q-learning, TD(0)-learning, TD(λ)-learning have been proposed. however, most of them have a drawback of very slow learning because the reinforcement value is given when they reach their goal state. In this thesis, we have proposed a reinforcement learning method that can approximate fast to the goal state in maze environments. The proposed reinforcement learning method is separated into global learning and local learning, and then it executes learning. Global learning is a learning that uses the replacing eligibility trace method to search the goal state. In local learning, it propagates the goal state value that has been searched through global learning to neighboring sates, and then searches goal state in neighboring states. we can show through experiments that the reinforcement learning method proposed in this thesis can find out an optimal solution faster than other reinforcement learning methods like Q-learning, TD(o)learning and TD(λ)-learning.

• Asian-Australasian Journal of Animal Sciences
• /
• 제18권9호
• /
• pp.1231-1236
• /
• 2005

Genetic parameters for first lactation milk production based on test day (TD) records of 56319 Iranian Holstein cows from 655 herds that first calved between 1991 and 2001 were estimated with restricted maximum likelihood method under an Animal model. Traits analyzed were milk yield and milk fat percentage. Heritability for TD records were highest in second half of the lactation, ranging from 0.11 to 0.19 for milk yield and 0.038 to 0.094 for milk fat percentage respectively. Estimates for lactation records for these traits were 0.24 and 0.26 respectively. Genetic correlations between individual TD records were high for consecutive TD records (>0.9) and decreased as the interval between tests increased. Estimates of genetic correlations of TD yield with corresponding lactation yield were highest (0.78 to 0.86) for mid-lactation (TD3 to TD8). Phenotypic correlations were lower than corresponding genetic correlations, but both followed the same pattern. For milk fat percentage no clear pattern was found. Results of this study suggested that TD yields especially in mid-lactation may be used for genetic evaluation instead of 305-day yield.