• 한국컴퓨터정보학회논문지
• /
• 제20권9호
• /
• pp.1-10
• /
• 2015

In this paper, we propose a new variable latency L1 data cache architecture for multi-core processors. Our proposed architecture extends the traditional variable latency cache to be geared toward the multi-core processors. We added a specialized data structure for recording the latency of the L1 data cache. Depending on the added latency to the L1 data cache, the value stored to the data structure is determined. It also tracks the remaining cycles of the L1 data cache which notifies data arrival to the reservation station in the core. As in the variable latency cache of the single-core architecture, our proposed architecture flexibly extends the cache access cycles considering process variation. The proposed cache architecture can reduce yield losses incurred by L1 cache access time failures to nearly 0%. Moreover, we quantitatively evaluate performance, power, energy consumption, power-delay product, and energy-delay product when increasing the number of cache access cycles.

• Journal of the Optical Society of Korea
• /
• 제16권2호
• /
• pp.107-114
• /
• 2012

In this paper, we propose a new modulation scheme called multi-coded variable pulse position modulation (MC-VPPM) for visible light communication systems. Two groups of signals (Pulse Width Modulation (PWM) and Pulse Position Modulation (PPM) groups) are multi-coded by orthogonal codes for transmitting data simultaneously. Then, each multi-level value of the multi-coded signal is converted to pulse width and position which results in not only an improved data rate, but also a processing gain in reception. In addition, we introduce average duty ratio and cyclic shift concepts in PWM through which dimming control for light illumination can be supported without any degradation in communication performance. Through simulation, we confirm that the proposed MC-VPPM shows a comparable BER curve and much greater achievable data rate than the conventional VPPM scheme using a visible light optical channel environment.

• 대한원격탐사학회지
• /
• 제22권3호
• /
• pp.211-219
• /
• 2006

A random forests classifier is applied to multi-sensor data fusion for supervised land-cover classification in order to account for the importance of variable. The random forests approach is a non-parametric ensemble classifier based on CART-like trees. The distinguished feature is that the importance of variable can be estimated by randomly permuting the variable of interest in all the out-of-bag samples for each classifier. Two different multi-sensor data sets for supervised classification were used to illustrate the applicability of random forests: one with optical and polarimetric SAR data and the other with multi-temporal Radarsat-l and ENVISAT ASAR data sets. From the experimental results, the random forests approach could extract important variables or bands for land-cover discrimination and showed reasonably good performance in terms of classification accuracy.

• Journal of the Korean Data and Information Science Society
• /
• 제27권2호
• /
• pp.499-510
• /
• 2016

생존분석 회귀모형에서 적절한 변수를 선택하는 것은 매우 중요하다. 본 논문에서는 "frailtyHL" R 패키지 (Ha 등, 2012)를 기반으로 하여 다수준 프레일티 모형 (multi-level frailty models)에서 벌점화 변수선택 방법 (penalized variable-selection method)의 절차를 소개한다. 여기서 모형 추정은 벌점화 다단계 가능도에 기초하며, 세 가지 벌점 함수 (LASSO, SCAD 및 HL)가 고려된다. 개발된 방법의 예증을 위해 벨기에 EORTC (European Organization for Research and Treatment of Cancer; 유럽 암 치료기구)에서 수행된 다국가/다기관 임상시험 자료를 이용하여 세 가지 변수 선택 방법의 결과를 비교하고, 그 결과들의 상대적 장 단점에 대해 토론한다. 특히, 자료 분석 결과에 의하면 SCAD와 HL방법이 LASSO보다 중요한 변수를 잘 선택하는 것으로 나타났다.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2005년도 Proceedings of ISRS 2005
• /
• pp.283-285
• /
• 2005

To account for the importance of variable in multi-sensor data fusion, random forests are applied to supervised land-cover classification. The random forests approach is a non-parametric ensemble classifier based on CART-like trees. Its distinguished feature is that the importance of variable can be estimated by randomly permuting the variable of interest in all the out-of-bag samples for each classifier. Supervised classification with a multi-sensor remote sensing data set including optical and polarimetric SAR data was carried out to illustrate the applicability of random forests. From the experimental result, the random forests approach could extract important variables or bands for land-cover discrimination and showed good performance, as compared with other non-parametric data fusion algorithms.

• 한국정보통신설비학회:학술대회논문집
• /
• 한국정보통신설비학회 2009년도 정보통신설비 학술대회
• /
• pp.382-386
• /
• 2009

Online or sequential learning is one of the most basic and powerful method to train neuron network, and it has been widely used in disease detection, weather prediction and other realistic classification problem. At present, there are many algorithms in this area, such as MRAN, GAP-RBFN, OS-ELM, SVM and SMC-RBF. Among them, SMC-RBF has the best performance; it has less number of hidden neurons, and best efficiency. However, all the existing algorithms use signal normal distribution as kernel function, which means the output of the kernel function is same at the different direction. In this paper, we use multi-variable normal distribution as kernel function, and derive EKF learning formulas for multi-variable normal distribution kernel function. From the result of the experience, we can deduct that the proposed method has better efficiency performance, and not sensitive to the data sequence.

• 대한산업공학회지
• /
• 제39권4호
• /
• pp.253-259
• /
• 2013

Investment scenarios in the transportation network design problem usually contain installation or expansion of multi-mode transportation links. When one applies the mode choice analysis and traffic assignment sequentially for each investment scenario, it is possible that the travel impedance used in the mode choice analysis is different from the user equilibrium cost of the traffic assignment step. Therefore, to estimate the travel impedance and mode choice accurately, one needs to develop a combined model for the mode choice and traffic assignment. In this paper, we derive the inverse demand and the excess demand functions for the multi-mode multinomial logit mode choice function and develop a combined model for the multi-mode variable demand traffic assignment problem. Using data from the regional O/D and network data provided by the KTDB, we compared the performance of the partial linearization algorithm with the Frank-Wolfe algorithm applied to the excess demand model and with the sequential heuristic procedures.

• 전기전자학회논문지
• /
• 제28권3호
• /
• pp.451-457
• /
• 2024

본 논문에서는 가격, 기후 요인, 수요, 수입량 등 다양한 변수를 데이터화한 후, LSTM(Long Short-Term Memory) 모델을 활용하여 농산물 가격을 예측하는 방법을 제안하였다. 시계열 데이터의 장기 의존성을 학습하는 LSTM 모델을 통해 예측 성능을 분석한 결과, 다양한 데이터를 통합함으로써 기존 방법보다 성능이 향상되었음을 확인하였다. 또한, 종속 변수인 가격 데이터 없이 독립 변수들만을 활용한 예측에서도 의미 있는 성과를 거두어, 모델의 발전 가능성을 확인할 수 있었다. 더 나아가, 다변수 모델을 사용할 경우 예측 성능이 더욱 개선될 수 있음을 알게 되었으며, 이러한 복합적인 접근이 배추 가격 예측의 정확도를 높이는 데 효과적임을 시사한다.

• Genomics & Informatics
• /
• 제16권4호
• /
• pp.33.1-33.9
• /
• 2018

Recently, there have been many studies in medicine related to genetic analysis. Many genetic studies have been performed to find genes associated with complex diseases. To find out how genes are related to disease, we need to understand not only the simple relationship of genotypes but also the way they are related to phenotype. Multi-block data, which is a summation form of variable sets, is used for enhancing the analysis of the relationships of different blocks. By identifying relationships through a multi-block data form, we can understand the association between the blocks in comprehending the correlation between them. Several statistical analysis methods have been developed to understand the relationship between multi-block data. In this paper, we will use generalized canonical correlation methodology to analyze multi-block data from the Korean Association Resource project, which has a combination of single nucleotide polymorphism blocks, phenotype blocks, and disease blocks.

• 대한공간정보학회지
• /
• 제18권4호
• /
• pp.101-107
• /
• 2010

본 연구의 목적은 벡터 도메인 안에 대규모 도로 선형 사상을 대상으로 실시간 데이터 변경, 관리가 가능한 네트워크의 다중 표현 데이터베이스 모델을 구축하는 것이다. 즉, 최상위 레벨의 네트워크 데이터로부터 이에 대응하는 하위 베이스 네트워크 데이터로 순차적으로 데이터 통합과 자동 매칭을 수행하는 상의하달 방식(top-down)을 기초로 하는 프레임워크를 제시하며, 이를 통해 변화 가능한 축척(variable-scale)의 지도를 생성하는 모델을 제안하였다. 구현된 MRDB(Multi-Resolution Database) 모델을 차량 항법 서비스에 적용하여 실제 동적 경로 안내 시스템에 활용 가능함을 확인할 수 있었다.