• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.2A
• /
• pp.85-90
• /
• 2009

In this paper, as an alternative to constant modulus algorithm based on MSE, maximization of the probability that equalizer output power is equal to the constant modulus of the transmitted symbols is introduced. The proposed algorithm using the gradient ascent method to the maximum probability criterion has superior convergence and steady-state MSE performance, and the error samples of the proposed algorithm exhibit more concentrated density functions in blind equalization environments. Simulation results indicate that the proposed training has a potential advantage versus MSE training for the constant modulus approach to blind equalization.

• Journal of architectural history
• /
• v.18 no.6
• /
• pp.127-144
• /
• 2009

This thesis aims to analyze the architectural form for the restoration of noogak(樓閣), Boje-roo(普濟樓) which was constructed at the central court of Beomeo-sa(梵魚寺). The results are as follow; 1. Boje-roo was established in 1700. By the constructional records and poetic essays of the times, the architectural of the establishment building was consisted fo two stories structure. The passage system for the approach to the court was the nuha-jinip(樓下進入), that is to say, one might walk between the columns of lower story of the building. And one could ascent the upper stairs and approach to the upper level, inner court. 2. By the remodelling in 1812, the building was varied to one story structure. Therefore the passage system for the approach to the inner court transformed from the nuha-jinip to the ugak-jinip(隅角進入), that is to say, one might ascent the atairs of the podium and could approach the inner court through the both sides of the building. 3. Under the Japanese imperialistic rule, the structure of it's roof and walls of Boje-roo were altered to the Japanese style. In 1965, the wooden columns were shifted to the concreted column, and the wooden wall-panels were shifted to the brick wall attached stone slates. 4. Conclusively speaking, for the restoration of Boje-roo, the establishment form in 1700 is the more proper than the remodelling form in 1812.

• Journal of the Korean Data and Information Science Society
• /
• v.7 no.1
• /
• pp.113-118
• /
• 1996

Principal component analysis(PCA) is an essential technique for data compression and feature extraction, and has been widely used in statistical data analysis, communication theory, pattern recognition, and image processing. Oja(1992) found that a linear neuron with constrained Hebbian learning rule can extract the principal component by using stochastic gradient ascent method. In practice real data often contain some outliers. These outliers will significantly deteriorate the performances of the PCA algorithms. In order to make PCA robust, Xu & Yuille(1995) applied statistical physics to the problem of robust principal component analysis(RPCA). Devlin et.al(1981) obtained principal components by using techniques such as M-estimation. The propose of this paper is to investigate from the statistical point of view how Xu & Yuille's(1995) RPCA works under the same simulation condition as in Devlin et.al(1981).

• Korean Journal of Applied Biomechanics
• /
• v.16 no.3
• /
• pp.173-181
• /
• 2006

The purpose of this study is grasp the problem of the gymnast, Kim, Dong-Hwa's Kip to Swallow Motion in Rings, and make up for the weak points to help him to perform a better performance. Therefore, two tryouts for $28^{th}$ Athens Olympic Games were filmed using video camera then finalized with Kinematical Analysis using 3D motion analysis program followings are the form of conclusions. 1. In the very first tryout, when he was doing a Swallow Support Scale, his CM position was high and arm slope was deduction because when he was doing Kip, the ascent velocity was low and he tried excessively to pull him on rings due to relying upon angular movement of shoulder joint. 2. When he was doing drop, he let his hip angle bend only little bit and let fall so making shoulder angle wider and maintain the level horizontally occurs strong drop motion when vertical descent is happening. 3. As a result, lowering the direction of a kick makes CM's movement path lower, increase vertical ascent velocity, and it helps to do the Swallow Support motion in short period of time. 4. After a strong drop motion, which is deep and fast, would make rope of ring shake so there is a defect that the body moves to forward area. However, it does not effect in Swallow Support Scale motion. 5. In the second tryout, trunk rotation angle and arm slope was fixed decrease while doing rotary motion. When rotary motion was happening, before the body was going under the rings, maintained his arm slope horizontally so his Swallow Support Scale motion was nearly perfect.

• Journal of Haehwa Medicine
• /
• v.8 no.2
• /
• pp.317-327
• /
• 2000

Through the study about Che-Qi-Jung(諸氣證) concerned with phenomenon of stress, we concluded as fallow. 1. Stressors are recognized three parts, internal cause, external cause, non-interexternal cause, in the oriental medicine. And the reaction of the body about stressors, was expressed change of Qi(氣). 2. The vascular system and autonomic nervous system are tensed or relexed. Because the etiology of Yuk-Em(六淫), Chil-Jeong(七情) make lose up and down, in and out, circulation of Gi(氣). The symtom caused with stress are explained, divided into Chil-Qi(七氣), Gu-Qi(九氣), Joong-Qi(中氣), Qi-Tong(氣痛), Qi-Yuk(氣逆), Qi-Wool(氣鬱). 3. The symtom of Chil-Qi is recognized into psycholoQical stress and Mae-Haek-Qi(梅核氣). The symtom of Gu-Qi is composed of Chil-Qi and Han(寒), Youl(熱), Ro(勞). 4. The symtoms of Qi-Yuk and Joong-Qi are caused by the ascent of Hwa-Qi(火氣). And Hwa-Qi is made by rage and depression. Qi-Yuk is chronic symtom that the ascent of Hwa-Qi is made by out of function in up and down mechanism of Qi. Joong-Qi is acute symtom that patholoQical state of Qi-Yuk take place for a while by sudden psycholoQical shock. 5. The symtom of Qi-Wool is loss of funtion made by depressed state in physioloQical mechanism, and Qi-Wool is observed in depressive psychosis and hypochondriasis and psychogenic. 6. The symtom of Qi-Tong is the pain caused by interference of circulation of Qi-Hyul(氣血). PsycholoQical stimulation such as rage and depression, tense autonomic nervous system. This make convulsion in gastro-intestine or vascular circuluation or abdominal muscles, and the pain is caused by convulsion.

• Korean Journal of Applied Biomechanics
• /
• v.25 no.3
• /
• pp.249-256
• /
• 2015

Objective : The purposes of this study was to perform a kinematical analysis on age and walkway types in elderly women subjects. Method : Forty subjects participated in the experiment (A1 group - age: $67.30{\pm}1.49yrs$, height: $153.81{\pm}4.47cm$, weight: $61.80{\pm}5.24kg$, A2 group - age: $71.70{\pm}1.10yrs$, height: $152.01{\pm}2.84cm$, weight: $59.69{\pm}7.34kg$, A3 group - age: $76.80{\pm}0.98yrs$, height: $150.16{\pm}6.08cm$, weight: $57.27{\pm}6.42kg$, A4 group - age: $81.80{\pm}0.60yrs$, height: $152.18{\pm}4.77cm$, weight: $55.80{\pm}7.78kg$). The study method adopted was the 3D analysis with six cameras. Ground type were classifed as gait pattern on flat, ascent and descent ramp. For the statistical analysis, the SPSS 21.0 was used to perform Repeated measured Two-way ANOVA. Results : In velocity of CM, there was faster movement on flat ground. When it came to the velocity of right toe, there was no significance in early mid-swing of right foot, but A4 was the slowest in late mid-swing of right foot on flat ground. In joint angle in left foot strike, the left hip joint and knee joint were more flexed in descent ramp, In addition left and right ankle joints were more plantarflexed in descent ramp, and left ankle joint was more plantarflexed in the over 75 yrs age groups. Conclusion : The higher age group were more flexed in lower body joints during descent ramp.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37A no.10
• /
• pp.865-875
• /
• 2012

In this paper, we optimize antenna locations for a distributed antenna system (DAS) with distributed antenna (DA) ports equipped with multiple antennas under per-DA port power constraint. Maximum ratio transmission and scaled zero-forcing beamforming are employed for single-user and multi-user DAS, respectively. Instead of maximizing the cell average ergodic sum rate, we focus on a lower bound of the expected signal-to-noise ratio (SNR) for the single-cell scenario and the expected signal-to-leakage ratio (SLR) for the two-cell scenario to determine antenna locations. For the single-cell case, optimization of the SNR criterion generates a closed form solution in comparison to conventional iterative algorithms. Also, a gradient ascent algorithm is proposed to solve the SLR criterion for the two-cell scenario. Simulation results show that DAS with antenna locations obtained from the proposed algorithms achieve capacity gains over traditional centralized antenna systems.

• 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.

• The Korean Journal of Ecology
• /
• v.8 no.1
• /
• pp.53-59
• /
• 1985

Phenological development, flowering and leafin times of Rhododendron muronulatum and R. schlippenbachii along elevagtion were studied at Mt. Kwanak, 629m high above the sea level, in Seoul. Flowering and leafing time of the former were delayed at the rate of 2.3~3.3 days and those of the latter were of 2.0~3.0 days per 100m ascent. Phenological changes of both plant species were closely correlated with minimum air temperature first, and then soil water content and minimum soil temperature among the climatic factors. Phenological difference caused by altitude and slope direction(southeast-northwest) among the topographic factors was admitted at the high significance level, but the difference by ridgevalley was little.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2007.04a
• /
• pp.33-36
• /
• 2007

최근에 인공지능 분야에서는, 국내외적으로 강화학습(reinforcement learning)에 관한 관심이 크게 증폭되고 있다. 강화학습의 최근 경향을 살펴보면, 크게 가치함수를 직접 활용하는 방법(value function-based methods), 제어 전략에 대한 탐색을 활용하는 방법(policy search methods), 그리고 액터-크리틱 방법(actor-critic methods)의 세가지 방향으로 발전하고 있음을 알 수 있다. 본 논문에서는 이중 세 번째 부류인 액터-크리틱 방법 중 NAC(natural actor-critic) 기법의 한 종류인 RLS-NAC(recursive least-squares based natural actor-critic) 알고리즘을 다양한 트레이스 감쇠계수를 사용하여 연속제어입력(real-valued control inputs)으로 제어되는 Kimura의 기는 로봇에 대해 적용해보고, 그 성능을 기존의 SGA(stochastic gradient ascent) 알고리즘을 이용하여 학습한 경우와 비교해보도록 한다.