• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.517-519
• /
• 2022

In 5G network environment, proactive mobility management is essential as 5G mobile networks provide new services with ultra-low latency through dense deployment of small cells. The importance of a system that actively controls device handover is emerging and it is essential to predict mobile trajectory during handover. Sequence-to-sequence model is a kind of deep learning model where it converts sequences from one domain to sequences in another domain, and mainly used in natural language processing. In this paper, we developed a system for predicting mobile trajectory in a wireless network environment using sequence-to-sequence model. Handover speed can be increased by utilize our sequence-to-sequence model in actual mobile network environment.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2020.11a
• /
• pp.168-171
• /
• 2020

초저지연성을 요구하는 5G 네트워크 환경에서 기기의 핸드오버를 능동적으로 조절하는 시스템의 중요성이 대두되고 있으며, 특히 핸드오버 시 기기의 이동성을 예측하는 것은 필수적이다. 딥러닝 모델의 일종인 생성적 적대 신경망은 두 신경망 사이의 경쟁 구도를 이용하여 두 신경망의 성능을 모두 높이는 목적으로 사용된다. 본 논문에서는 주로 데이터 생성 모델로 사용되는 생성적 적대 신경망을 이용하여 무선 네트워크 환경에서 기기의 이동성을 예측하는 시스템을 개발하였다. 이를 통해 실제 모바일 네트워크 환경에 적용되었을 경우 핸드오버 속도를 높이도록 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.633-636
• /
• 2021

4G 에서 5G 로 기술이 발전하며 무선 통신에 필요한 자원이 급격히 증가하고 있다. 증가된 자원을 효율적으로 관리하는 것은 필수적이며 이를 위해 paging cost 감소 연구들이 진행되고 있다. 순환신경망을 응용한 paging cost 감소 연구에서는 연속 예측으로 인해 예측 정확도 감소 문제가 발생한다. 본 논문에서는 강화학습 기반 이동성 예측 기법을 제안하고 기존 순환신경망 응용 기법에서 발생하는 정확도 감소 문제를 극복한다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.37-38
• /
• 2022

농산물의 산지 가격이나 도매가격이 등락하면, 즉시 또는 일정한 시차 이후에 소비자가격도 등락한다. 본 논문에서는 선형회귀모델을 통해 쌀 가격을 예측하고 쌀 가격에 영향을 미치는 날씨의 시기를 찾아보고자 한다. 이에 따라 KAMIS, 기상자료개방포털, KOSIS에서 수집한 날씨, 생산량, 그리고 소비자물가 등락률 데이터를 이용하여 쌀 가격 예측을 수행하고, 날씨 데이터와 쌀 가격 데이터의 날짜 간격을 두어 날씨가 쌀 가격에 영향을 미치는 시기를 알아보았다. 모델 평가 결과, 2개월 간격을 두고 예측한 RMSE가 164.135로 가장 큰 영향을 미쳤다. 본 연구를 기반으로 향후 다른 농산물의 가격 예측도 가능할 것이며 농산물에 영향을 미치는 변수의 시기도 예측할 수 있을 것으로 기대한다.

• The Journal of the Korea Contents Association
• /
• v.18 no.5
• /
• pp.265-279
• /
• 2018

This study conducted an experiment using data mining techniques to develop prediction models of worker job turnover. The experiment used data from the '2015 Graduate Occupational Mobility Survey' by the Korea Employment Information Service. We developed the prediction models using a decision tree, Bayes net, and artificial neural network. We found that the decision tree-based prediction model reported the best accuracy. We also found that the six influential factors affecting employees' turnover intention are type of working time, job status, full-time or not full-time, regular working hours per week, regular working days per week, and personal development opportunities. From the decision tree-based prediction model, we derived 12 rules of employee turnover for all job types. Using the derived rules, we proposed helpful directions for enhancing workers' job tenure. In addition, we analyzed the influential factors affecting employees' job turnover intention according to four job types and derived rules for each: office (ten rules), culture and art (nine rules), construction (four rules), and information technology (six rules). Using the derived rules, we proposed customized directions for improving the job tenure for each group.

• Journal of Soil and Groundwater Environment
• /
• v.20 no.6
• /
• pp.133-139
• /
• 2015

Attention to munitions constituents such as 2,4,6-trinitrotoluene (TNT) and hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) in the firing ranges is increasing due to their toxicity and high mobility to the environment. It is helpful to use a systemic model to predict the amount of contaminants for the establishment of environmental management of firing ranges. This study employed Training Range Environmental Evaluation and Characterization System (TREECS) program to estimate the mobility characteristics of TNT and RDX via groundwater leaching, soil erosion and surface water runoff. The prediction results of the TNT and RDX migration with TREECS showed that 68% of initial TNT and 21% of initial RDX were discharged through the soil erosion and the 20% of initial TNT and 54% of initial RDX ran out the firing range via the groundwater leaching. The rest of the initial TNT and RDX moved to adjacent surface water via surface runoff. The data suggest that soil erosion and surface runoff occupying 80% of TNT to the total amount are important migration pathways. On the other hand, groundwater leachning occupying 54% to the total amount was also important pathway for RDX.

• Membrane Journal
• /
• v.3 no.2
• /
• pp.79-82
• /
• 1993

The sorption equilibria and permeabilities for $CO_2$ in a homogeneous membrane of polystyrene with the glass transition temperature of $95^{\circ}C$ were measured at a temperature of $60^{\circ}C$ and gas pressures up to 1.6 MPa and 2.5 MPa, respectively. The sorption isotherm had the form af dual-mode sorption model at low gas pressures, but became linear at pressures above 1.3 MPa. The linear portion of the isotherm extrapolated to the origin. The pressure dependence of the rnean permeability coefficient deviated upward from the dualsrhode mobility model prediction. It was found that the glass transition was brought out by the plasticization action of sotbed $CO_2$ at a gas pressure of 1.3 MPa from the sorption isotherm. And this result was consistent with an increase in the mean permeability coefficient with applied gas pressure.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.4A
• /
• pp.370-377
• /
• 2004

In this paper, we present an adaptive modulation technique for orthogonal frequency division multiplexing (OFDM) for broadband wireless communications. Also, using improved channel prediction, we enhance the performance of adaptive OFDM in high mobility environments. Adaptive modulation technique has been shown to achieve reliable high-rate data transmission over frequency-selective fading channel when OFDM is employed. This scheme requires the accurate channel information between two stations for a better performance. In an outdoor high mobility environment, most of adaptive OFDM systems have to be given the channel information transmitted from the receiver. Even if it is possible, there is some delay. Moreover, the channel impulse response between two stations is very rapidly varied. If the channel information is obsolete at the time of transmission, then poor system performance will result. In order to solve this problem, we propose adaptive OFDM with improved channel predictor. The proposed bit allocation algorithm has a lower complexity and the proposed scheme mitigates the effect of channel delay. Robust approach is less sensitive to outdated channel information. Performance results show that the proposed scheme can achieve considerable performance enhancement.

• Journal of Korean Society of Transportation
• /
• v.31 no.5
• /
• pp.79-88
• /
• 2013

Recently, due to THE MOBILITY ENHANCEMENT FOR THE MOBILITY IMPAIRED ACT, local governments have tired to make various efforts on special transport services(STS), low-flow bus, and installing elevator in subway stations for handicapped people. But in case of STS, insufficient numbers of taxi are raised against the increasing demand of hadicapped people due to the limited budget. This study investigated actual use condition of STS and characteristics of selection of handicapped people on Duribal. In addition, an ordered-logistic model was employed for developing taxi use prediction model considering taxi fare discounts for diverting Duribal demands to taxies. The results can be a significant basic data for transportation policies to improve travel efficiency of the handicapped.

• Korean Journal of Environmental Agriculture
• /
• v.24 no.4
• /
• pp.350-357
• /
• 2005

This study was conducted to investigate the downward mobility of pesticides using soil columns and to compare the experimental results with values predicted from Convective mobility test model. Nine pesticides such as metolcarb, molinate, fanobucarb, isazofos, diazinon, fenitrothion, dimepiperate, parathion and chlorpyrifos-methyl were used for leaching test in soil column for four soils; Jungdong (upland soil), Gangseo (paddy soil), Yesan (forest soil), and Sineom(upland, volcanic ash-derived soil) series. The peak concentrations leached from 10 cm-columns of three soils except Sineom series ranged 6.5 to 12.6 mg/L for metolcarb, 2.6 to 5.0 mg/L for molinate, 4.5 to 7.8 mg/L for fenobucarb, 0.39 to 1.36 mg/L for dimepiperate, 1.1 to 4.6 mg/L for isazofos, 0.01 to 0.14 mg/L for diazinon, lower than 0.01 to 0.70 mg/L for fenitrothion and lower than 0.01 to 0.44 mg/L for parathion. But chlorpyrifos-methyl was not leached from any soil columns. Elution volumes to reach the peak of metolcarb, molinate, fenobucarb, isazofos, diazinon, and dimepiperate in the leachate ranged 1.1 to 2.1 pore volume (PV), 1.6 to 3.3 PV, 1.6 to 3.3 PV, 2.1 to 4.4 PV, 6 to 15 PV, and 8 to 21 PV, respectively. On the same water flux conditions, convection times estimated by Convective mobility test model were coincided with results from soil column test in most of the soil-pesticide combinations applied. Based on convection times estimated by the model at standard conditions (water flux 1 cm/day), metolcarb was classified as most mobile, molinate, fenobucarb and isazofos as mobile or most mobile, dimepiperate as moderately mobile or mobile, diazinon as mobile, fenitrothion and parathion as slightly mobile or mobile and chlorpyrifos-methyl as immobile or slightly mobile.