• Journal of the Korean Data and Information Science Society
• /
• 제24권2호
• /
• pp.321-332
• /
• 2013

본 연구는 지속적인 저출산의 여파로 2020년에는 초등학생 수가 올해 대비 17%, 중고교생은 30%가 감소할 것이라는 예측을 가지고 초등학교 교원 수를 예측하기 위한 방법을 제시하는데 있다. 교육통계연보의 1970년부터 2010년까지의 초등교육 관련 주요 통계 자료를 이용하여 시계열 회귀모형과 시계열 그룹별 회귀모형, 지수평활법 모형을 제시하고, 제시된 모형을 이용하여 향후 10년간의 연도별 초등학교 교원 수를 예측하였다. 모형 예측 결과 시계열 그룹별 회귀 모형이 교원 수 시계열을 가장 잘 설명하는 것으로 나타났으며, 적합한 모형으로 판명되었다. 3가지 분석방법 모형에 따른 예측값에 대한 장단점과 한계를 제시한다.

• 응용통계연구
• /
• 제30권1호
• /
• pp.181-193
• /
• 2017

본 논문에서는 계층적 시계열 자료 분석을 위한 대표적인 두 가지 방법인 상향식과 최적조합 예측법을 소개한다. 이러한 예측법은 계층적 시계열을 구성하는 모든 계열을 예측해야 하는 독립적 예측과 달리, 임의의 조정 과정이 없이 하위 계층 계열의 예측값의 합은 항상 상위 계층의 예측값과 일치하게 된다. 또한, 독립적 예측과 비교하여 예측력을 향상시킨다. 계층적 예측법의 효율성을 살펴보기 위하여 국내 16개 시도별 남녀 교통사고 발생건수 시계열 자료를 예측하였다. 이를 통하여 교통사고 발생건수에 대한 각 계층의 예측에서 계층적 방법과 독립적 방법의 차이점 및 우수성을 비교하였다.

• Communications for Statistical Applications and Methods
• /
• 제29권6호
• /
• pp.695-708
• /
• 2022

Recently, several studies have been conducted using state space model. In this study, a dynamic linear model with state space model form is applied to stock data. The monthly returns for 135 Korean stocks are fitted to a dynamic linear model, to obtain an estimate of the time-varying 𝛽-coefficient time-series. The model formula used for the return is a capital asset pricing model formula explained in economics. In particular, the transition equation of the state space model form is appropriately modified to satisfy the assumptions of the error term. k-shape clustering is performed to classify the 135 estimated 𝛽 time-series into several groups. As a result of the clustering, four clusters are obtained, each consisting of approximately 30 stocks. It is found that the distribution is different for each group, so that it is well grouped to have its own characteristics. In addition, a common pattern is observed for each group, which could be interpreted appropriately.

• 대한치과교정학회지
• /
• 제47권5호
• /
• pp.298-305
• /
• 2017

Objective: This paper describes changes in the characteristics of patients seeking orthodontic treatment over the past decade and the treatment they received, to identify any seasonal variations or trends. Methods: This single-center retrospective cohort study included all patients who presented to Seoul National University Dental Hospital for orthodontic diagnosis and treatment between January 1, 2005 and December 31, 2015. The study analyzed a set of heterogeneous variables grouped into the following categories: demographic (age, gender, and address), clinical (Angle Classification, anomaly, mode of orthodontic treatment, removable appliances for Phase 1 treatment, fixed appliances for Phase 2 treatment, orthognathic surgery, extraction, mini-plate, mini-implant, and patient transfer) and time-related variables (date of first visit and orthodontic treatment time). Time series analysis was applied to each variable. Results: The sample included 14,510 patients with a median age of 19.5 years. The number of patients and their ages demonstrated a clear seasonal variation, which peaked in the summer and winter. Increasing trends were observed for the proportion of male patients, use of non-extraction treatment modality, use of ceramic brackets, patients from provinces outside the Seoul region at large, patients transferred from private practitioners, and patients who underwent orthognathic surgery performed by university surgeons. Decreasing trends included the use of metal brackets and orthodontic treatment time. Conclusions: Time series analysis revealed a seasonal variation in some characteristics, and several variables showed changing trends over the past decade.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권5호
• /
• pp.2539-2554
• /
• 2017

Regression-based image super resolution (SR) methods have shown great advantage in time consumption while maintaining similar or improved quality performance compared to other learning-based methods. In this paper, we propose a novel single image SR method based on hierarchical regression to further improve the quality performance. As an improvement to other regression-based methods, we introduce a hierarchical scheme into the process of learning multiple regressors. First, training samples are grouped into different clusters according to their geometry similarity, which generates the structure layer. Then in each cluster, a compact dictionary can be learned by Sparse Coding (SC) method and the training samples can be further grouped by dictionary atoms to form the detail layer. Last, a series of projection matrixes, which anchored to dictionary atoms, can be learned by linear regression. Experiment results show that hierarchical scheme can lead to regression that is more precise. Our method achieves superior high quality results compared with several state-of-the-art methods.

• 한국초지조사료학회지
• /
• 제17권3호
• /
• pp.305-314
• /
• 1997

Nineteen clones of Jerusalem Artichoke (JA) from several countries were collected through the series of experiments about JA started in 1979. Collected clones were screened for adaptibility in Korea and showed introduction path way. The results about an ecological response of collected clones including flowering, tuberization, biomass production, sugar contents and grouping of collected clones for use of genetic material were as follows; 1. Nineteen clones collected were ffom Korea(2), Japan(l), USA(Z), Canada(2), France(4), Germany(7), and USSR(1). 2. Through the characteristics of top collected clones were divided to the types of branch-non branch, short and long plant height, and early and late maturity. Tuber characteristics were mainly grouped to the types of white skin color-violet skin color, clusters-single unit, round-elongate, and knotty-smooth. 3. Total sugar yields 6-om top at flowering time were 490 - 630kgl10a and 6-om the tuber were 420 -490 kg/ IOa through the high yielding clones. The top-high yielding clones were Mammoth French White, Fuseau 60, Nahodka, and JA3. The higher tuber yields were got from the clones of D- 19, Colombia, Bianka and Mammoth French White. 4. Collected clones were grouped to three and first group was characterized to early maturity and short plant height and second group to medium and finally, third group to late maturity and high plant height. 5. High yielding of top was 6-om the I group of early maturity and short plant height and high yielding of tuber h m III group of late maturity and high plant height.

• 기술혁신학회지
• /
• 제7권3호
• /
• pp.681-709
• /
• 2004

The purpose of this study is to classify customers by e-mail responsiveness on time-series analysis and testify the effectiveness of grouping by ROI analysis. Response recency, response frequency and Activity(RFA) of e-mailing systems are adapted for Customer segmentations. ROI analysis are consisted of open, click-through, duration time, personalization, conversion rate and email loyalty index of email systems. Major findings are as follows: RFA analysis is used for customer segmentations that is fundamental process of e-CRM applications. Customers can be grouped into loyal customers, odds customers, dormant customers, secession customers, and observation customers by RFA grouping. Loyal customer group has high point in all ROI index compared to other groups. These results indicated that customer responsiveness of e-mail systems were appropriate methods to group the customer with demographic variables. Therefore, effective e-mail marketing strategy of e-Biz should have suitable active DB and Behavior targeting is best approach to enforce the target e-mail marketing.

• Journal of the korean society of oceanography
• /
• 제32권1호
• /
• pp.38-45
• /
• 1997

To find out the changes in vertical distribution patterns over the 24-h period, a key and the first step to tackle the problem of adaptive significance of diel vertical migration (DVM), vertically stratified time series samplings with multiple opening/closing plankton samplers were done in the East Sea of Korea (Sea of Japan). Sampling was done almost every 4 h for one day period following the same water parcel in Nov. 1995 and May 1996, respectively. Resultant patterns of vertical distribution showed that some species such as most abundant taxa Metridia pacifica and Scolecithyicella minor, both Copepoda, performed DVM even in the study area of strong thermal stratification. Their patterns of DVM such as distance scales and timing of movements were not the same each other, and they were separated from other taxa in the dendrogram obtained by the cluster analyses, Most minor taxa grouped in one, however, seemed not to do DVM in the study area of strong thermal stratification. They usually preferred the warmer surface layer where the foods were probably more abundant.

• 수산해양기술연구
• /
• 제53권4호
• /
• pp.386-395
• /
• 2017

For scientific research, a number of acoustic surveys using commercial echosounders equipped in fishing vessels were conducted throughout the world; however, few studies were performed in South Korea. Hence, this research is an preliminary study for presenting the application of a sounder from a fishing vessel. The fishing operations using a pair trawler (7 Cheonghae) was conducted in the Northwest-Western sea of Jeju Island from 20 to 23 April, 2016. Substantial impulse noises and attenuated signals were eliminated by the latest algorithms. Acoustic signals were grouped into the fish aggregations and long layer-like signals. The fish aggregations appeared between 30 and 60 m, and long layer-like signals showed the diurnal vertical migration. Energetic, morphological and positional properties of the fish aggregations and layer-like signals were described. The fish aggregations appeared mainly between sunrise and sunset; however layer-like signals tended to be presented regardless of time in consideration of the time series analysis. On the basis of the consignment sales, Scomberomorus niphonius, the target species of F/V 7 Cheonghae, was the highest catch with 4,280 kg (74.6%) and might have appeared in fish aggregations and layer forms.

• 응용통계연구
• /
• 제29권4호
• /
• pp.581-594
• /
• 2016

본 논문은 이노베이션 상태공간모형을 근간으로 기존의 지수평활법을 포괄할 수 있는 다중 계절형 모형을 소개한다. 특히 이 모형은, 기존 모형의 한계를 극복하고 동일한 계절 내의 다양성을 표현할 수 있도록 계절 성분을 행렬로 표현하는 정교한 구조를 가지고 있다. 이런 구조를 이용하면 비슷한 패턴을 가지는 계절 성분의 모수를 그룹별로 분류할 수 있다. 따라서, 다중 계절형 모형은 모수절약 원칙을 달성할 수 있으며 모형의 해석이 용이한 장점을 가지고 있을 뿐만 아니라, 잠재적으로 임의의 개수의 계절성도 수용 가능하다. 본 연구에서는 다중 계절형 모형을 이용하여 시간 단위로 관측된 한국 전력 수요량을 분석하고 예측한다. 특히, 시간별 전력 수요량의 계절성은 1일 및 1주일의 두 가지로 고려되었고 이를 토대로 유사한 요일들은 공통 계절로 그룹화하였다. 모형의 예측 성능을 평가하기 위하여 기존 지수평활법의 예측 결과와 비교하였다. 그 결과, 다중 계절형 모형이 기존 지수평활법보다 예측력이 우수함을 확인하였다.