• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2020년도 추계학술발표대회
• /
• pp.20-23
• /
• 2020

This research is to design an effective prefetching method required for DRAM-PCM hybrid main memory systems especially used for big data applications and massive-scale computing environment. Conventional prefetchers perform well with regular memory access patterns. However, workloads such as graph processing show extremely irregular memory access characteristics and thus could not be prefetched accurately. Therefore, this research proposes an efficient dynamical prefetching algorithm based on the regression method. We have designed an intelligent prefetch engine that can identify the characteristics of the memory access sequences. It can perform regular, linear regression or polynomial regression predictive analysis based on the memory access sequences' characteristics, and dynamically determine the number of pages required for prefetching. Besides, we also present a DRAM-PCM hybrid memory structure, which can reduce the energy cost and solve the conventional DRAM memory system's thermal problem. Experiment result shows that the performance has increased by 40%, compared with the conventional DRAM memory structure.

• Communications for Statistical Applications and Methods
• /
• 제9권1호
• /
• pp.229-240
• /
• 2002

Whether to use linear or quadratic model in the analysis of regression data is one of the important problems in classical regression model and measurement error model (MEM). In MEM, four goodness of fit test statistics are available In solving that problem. Two are from the derivation of estimators of quadratic MEM, and one is from that of the general $k^{th}$-order polynomial MEM. The fourth one is derived as a variation of goodness of fit test statistic used in linear MEM. The purpose of this paper is to find the most powerful test statistic among them through the small-scale simulation.

• Communications for Statistical Applications and Methods
• /
• 제19권2호
• /
• pp.293-301
• /
• 2012

Quantile regression proposed by Koenker and Bassett (1978) is a statistical technique that estimates conditional quantiles. The advantage of using quantile regression is the robustness in response to large outliers compared to ordinary least squares(OLS) regression. A regression tree approach has been applied to OLS problems to fit flexible models. Loh (2002) proposed the GUIDE algorithm that has a negligible selection bias and relatively low computational cost. Quantile regression can be regarded as an analogue of OLS, therefore it can also be applied to GUIDE regression tree method. Chaudhuri and Loh (2002) proposed a nonparametric quantile regression method that blends key features of piecewise polynomial quantile regression and tree-structured regression based on adaptive recursive partitioning. Lee and Lee (2006) investigated wage determinants in the Korean labor market using the Korean Labor and Income Panel Study(KLIPS). Following Lee and Lee, we fit three kinds of quantile regression tree models to KLIPS data with respect to the quantiles, 0.05, 0.2, 0.5, 0.8, and 0.95. Among the three models, multiple linear piecewise quantile regression model forms the shortest tree structure, while the piecewise constant quantile regression model has a deeper tree structure with more terminal nodes in general. Age, gender, marriage status, and education seem to be the determinants of the wage level throughout the quantiles; in addition, education experience appears as the important determinant of the wage level in the highly paid group.

• Journal of the Korean Statistical Society
• /
• 제33권4호
• /
• pp.367-379
• /
• 2004

This article deals with the nonparametric analysis of longitudinal data when there exist possible correlations among repeated measurements for a given subject. We consider a quasi-likelihood regression model where a transformation of the regression function through a link function is linear in time-varying coefficients. We investigate the local polynomial approach to estimate the time-varying coefficients, and derive the asymptotic distribution of the estimators in this quasi-likelihood context. A real data set is analyzed as an illustrative example.

• 한국컴퓨터산업학회논문지
• /
• 제2권6호
• /
• pp.859-866
• /
• 2001

성공적인 프로젝트 계획은 활용 가능한 일정과 더불어 프로젝트를 완수하는데 요구되는 노력을 얼마나 정확히 추정하느냐에 달려있다. 새로운 또는 보다 나은 모델 개발에 많은 연구가 이루어졌지만 현존하는 소프트웨어 노력 추정 모델들은 개발 전순기에 대해 투입되는 총 개발노력과 단위시간당 소요되는 인력인 노력 함수만을 제공한다. 또한, Putnam은 세부단계별로 일정한 개발노력 투입 비율을 제시하였다. 그러나 소프트웨어의 규모, 복잡도와 운영환경의 영향으로 인해 프로젝트 별로 투입되는 총 개발노력의 규모에 차이가 발생하며, 그 결과, 개발 세부단계별로 투입되는 노력의 규모도 프로젝트마다 차이가 발생한다. 본 논문은 총 개발노력 변동에 따른 소프트웨어의 명세화, 구축과 시험단계에 투입될 개발노력을 추정하는 선형과 다항식 모델을 제시하였다. 이 모델들은 128개의 다른 소프트웨어 프로젝트들로부터 유도되었다. 제안된 모델은 프로젝트의 일정과 노력 할당 관리에 실질적인 지침을 제공할 것이다.

• 대한기계학회논문집A
• /
• 제21권10호
• /
• pp.1718-1729
• /
• 1997

This paper describes motion errors due to acceleration and deceleration types of servo motors in NC machine tools. Motion errors are composed of two components : one is due to transient response of a servomechanism and the other comes from gain mismatching of positioning servo motors. It deals with circular interpolation to identify motion errors by using Interface card. Also in order to minimize motion errors, this study presents an effective method to optimize parameters which are connected with motion errors. The proposed method is based upon a second order polynomial regression model and it includes an orthogonal array method to make the effective results of experiments. The validity and reliability of the study were verified on a vertical machining center equipped with FANUC 0MC through a series of experiments and analysis.

• 전자공학회논문지
• /
• 제52권12호
• /
• pp.109-117
• /
• 2015

본 논문에서는 DSP(Digital Signal Processor)를 이용하여 LED(Light Emitting Diode)에 전류 또는 전압을 공급하고, 이에 따라 나타나는 전압 또는 전류 특성을 분석하는 시스템에서, 전원 공급 또는 측정하는 회로의 비선형 오차 및 임의로 발생하는 오차를 감소시키는 방법을 제안하였다. 임의 오차를 줄이기 위해서는 재귀 평균 방법을 이용하였으며, 비선형 오차를 줄이기 위해서는 보정과정에서 획득한 데이터들을 2차 다항 회귀분석 방법을 이용하여 보정계수를 구하였으며, 이를 이용하여 LED를 생산 시 측정하는 항목인 역방향전류(IR), 역방향 전압(VR), 순방향전압(VF1, VF2, VF3)에 적용하여 오차를 교정하였다. 실험 결과에서는 오차율이 0.017 ~ 0.043 %로 관찰되었다.

• 대한조선학회논문집
• /
• 제42권5호
• /
• pp.534-541
• /
• 2005

Until now, Korean shipyards have accumulated a great amount of data. But they do not have appropriate tools to utilize the data in practical works. Engineering data contains experts' experience and know-how in its own. It is very useful to extract knowledge or information from the accumulated existing data by using data mining technique This paper treats an evolutionary computation based on genetic programming (GP), which can be one of the components to realize data mining. The paper deals with linear models of GP for the regression or approximation problem when given learning samples are not sufficient. The linear model, which is a function of unknown parameters, is built through extracting all possible base functions from the standard GP tree by utilizing the symbolic processing algorithm. In addition to a standard linear model consisting of mathematic functions, one variant form of a linear model, which can be built using low order Taylor series and can be converted into the standard form of a polynomial, is considered in this paper. The suggested model can be utilized as a designing tool to predict design parameters with small accumulated data.

• 농업과학연구
• /
• 제41권3호
• /
• pp.169-175
• /
• 2014

Objective of this study was to investigate effects of pre-processing method and number of sampling leaves on stability of the reflectance measurement for Chinese cabbage and kale leaves. Chinese cabbage and kale were transplanted and cultivated in a plant factory. Leaf samples of the kale and cabbage were collected at 4 weeks after transplanting of the seedlings. Spectra data were collected with an UV/VIS/NIR spectrometer in the wavelength region from 190 to 1130 nm. All leaves (mature and young leaves) were measured on 9 and 12 points in the blade part in the upper area for kale and cabbage leaves, respectively. To reduce the spectral noise, the raw spectral data were preprocessed by different methods: i) moving average, ii) Savitzky-Golay filter, iii) local regression using weighted linear least squares and a $1^{st}$ degree polynomial model (lowess), iv) local regression using weighted linear least squares and a $2^{nd}$ degree polynomial model (loess), v) a robust version of 'lowess', vi) a robust version of 'loess', with 7, 11, 15 smoothing points. Effects of number of sampling leaves were investigated by reflectance difference (RD) and cross-correlation (CC) methods. Results indicated that the contribution of the spectral data collected at 4 sampling leaves were good for both of the crops for reflectance measurement that does not change stability of measurement much. Furthermore, moving average method with 11 smoothing points was believed to provide reliable pre-processed data for further analysis.

• 품질경영학회지
• /
• 제23권3호
• /
• pp.84-92
• /
• 1995

A method is proposed for the simultaneous optimization of several response functions that depend on the same set of controllable variables and are adequately represented by a response surface model (polynomial regression model) with the same degree and with constraint that the individual responses have the target values. First, the multiple responses data are checked for linear dependencies among the responses by eigenvalue analysis. Thus a set of responses with no linear functional relationships is used in developing a function that measures the distance estimated responses from the target values. We choose the optimal condition that minimizes this measure. Also, under the different degree of importance two step procedures are proposed.