Why Gabor Frames? Two Fundamental Measures of Coherence and Their Role in Model Selection

Bajwa, Waheed U.,Calderbank, Robert,Jafarpour, Sina The Korea Institute of Information and Commucation 2010 Journal of communications and networks Vol.12 No.4

The problem of model selection arises in a number of contexts, such as subset selection in linear regression, estimation of structures in graphical models, and signal denoising. This paper studies non-asymptotic model selection for the general case of arbitrary (random or deterministic) design matrices and arbitrary nonzero entries of the signal. In this regard, it generalizes the notion of incoherence in the existing literature on model selection and introduces two fundamental measures of coherence-termed as the worst-case coherence and the average coherence-among the columns of a design matrix. It utilizes these two measures of coherence to provide an in-depth analysis of a simple, model-order agnostic one-step thresholding (OST) algorithm for model selection and proves that OST is feasible for exact as well as partial model selection as long as the design matrix obeys an easily verifiable property, which is termed as the coherence property. One of the key insights offered by the ensuing analysis in this regard is that OST can successfully carry out model selection even when methods based on convex optimization such as the lasso fail due to the rank deficiency of the submatrices of the design matrix. In addition, the paper establishes that if the design matrix has reasonably small worst-case and average coherence then OST performs near-optimally when either (i) the energy of any nonzero entry of the signal is close to the average signal energy per nonzero entry or (ii) the signal-to-noise ratio in the measurement system is not too high. Finally, two other key contributions of the paper are that (i) it provides bounds on the average coherence of Gaussian matrices and Gabor frames, and (ii) it extends the results on model selection using OST to low-complexity, model-order agnostic recovery of sparse signals with arbitrary nonzero entries. In particular, this part of the analysis in the paper implies that an Alltop Gabor frame together with OST can successfully carry out model selection and recovery of sparse signals irrespective of the phases of the nonzero entries even if the number of nonzero entries scales almost linearly with the number of rows of the Alltop Gabor frame.

Why Gabor Frames? Two Fundamental Measures of Coherence and Their Role in Model Selection (Invited Paper)

Waheed U. Bajwa,Robert Calderbank,Sina Jafarpour 한국통신학회 2010 Journal of communications and networks Vol.12 No.4

The problem of model selection arises in a number of contexts, such as subset selection in linear regression, estimation of structures in graphical models, and signal denoising. This paper studies non-asymptotic model selection for the general case of arbitrary (random or deterministic) design matrices and arbitrary nonzero entries of the signal. In this regard, it generalizes the notion of incoherence in the existing literature on model selection and introduces two fundamental measures of coherence—termed as the worst-case coherence and the average coherence—among the columns of a design matrix. It utilizes these two measures of coherence to provide an in-depth analysis of a simple, model-order agnostic one-step thresholding (OST) algorithm for model selection and proves that OST is feasible for exact as well as partial model selection as long as the design matrix obeys an easily verifiable property, which is termed as the coherence property. One of the key insights offered by the ensuing analysis in this regard is that OST can successfully carry out model selection even when methods based on convex optimization such as the lasso fail due to the rank deficiency of the submatrices of the design matrix. In addition, the paper establishes that if the design matrix has reasonably small worst-case and average coherence then OST performs near-optimally when either (i) the energy of any nonzero entry of the signal is close to the average signal energy per nonzero entry or (ii) the signal-to-noise ratio in the measurement system is not too high. Finally, two other key contributions of the paper are that (i)it provides bounds on the average coherence of Gaussian matrices and Gabor frames, and (ii) it extends the results on model selection using OST to low-complexity, model-order agnostic recovery of sparse signals with arbitrary nonzero entries. In particular, this part of the analysis in the paper implies that an AlltopGabor frame together with OST can successfully carry out model selection and recovery of sparse signals irrespective of the phases of the nonzero entries even if the number of nonzero entries scales almost linearly with the number of rows of the Alltop Gabor frame.

A Study of Model Selection for Electric Data using Cross Validation Approach

Saraswathi Sivamani,Saravana Kumar,신창선,박장우,조용윤 한국지식정보기술학회 2017 한국지식정보기술학회 논문지 Vol.12 No.6

In this paper, the appropriate model is selected for the risk assessment of the electric utility pole data with the help of cheat sheets and k-fold cross validation. In order to analyze, predict and forecast the data, the appropriate model has to be selected. The major issue is the declination of the accuracy in the model fitting, which may result in poor model selection. There are different type of machine learning algorithm, which makes it difficult to conclude the model selection. To ensure the proper selection of the model, we undergo a two-step process. Firstly, the basic model is selected with the existing model selection cheat sheets named as Scikit learn and Microsoft azure, by understanding the available input and required output of the data. After getting through the multiple question, the respective models such as Generalized Additive Model, Generalized Linear Model, Linear Regression and Support Vector Machine are obtained. In order to attain the appropriate model, we perform k-fold cross validation to estimate the risk of the algorithms, by comparing 2-fold, 8-fold and 10-fold cross validation. Between the three set, the 10-cross fold validation of generalized additive model is selected with the least risk error. Using k-fold cross validation, we estimate the accuracy of the model that is suitable for the data, by using the electric power data set.

고등학생의 대학전공선택 프로그램 모형 개발

이건남,정철영 한국농·산업교육학회 2009 농업교육과 인적자원개발 Vol.41 No.1

This study was to develop college major selection program model, to identify content factors for each step, and to test validity of finalized college major selection program model. To pursue this goal, the concept of a college major selection program was developed through a literature review. The characteristics of college major selections by high school students were identified through an analysis of survey results regarding the current situation and characteristics of high school students in their first year. These included the procedure of model development, the task of the career development of high school students, the career development ability of high school students, and the college major selection program. An initial program model composed of six steps, 16 categories, and 36 sub-divisions of the program model and content factors was identified. Delphi was conducted to assess the validity of the identified program model and initial content factors. A panel was included that consisted of five expert professors, researchers, and practitioners who operate programs related to college entrance, as well as career counseling teachers for a total of 15 panels from each section. With the Delphi method, the validity of the draft was confirmed by two rounds. Data from each rounds of Delphi surveys were analyzed for means, standard deviations, degree of coincidence and convergence in opinion, agreement rate, ICC(1), ICC(2) and degree of consensus by SPSS 15.0 version and EXCEL. Whether Delphi survey would continue or not an additional round, was determined by agreement rate(80% or more), CVR(≻0.49 in case of 15 experts), degree of consensus(standard deviations<0.800), ICC(1)(≧0.05), ICC(2)(≺0.70). The college major selection program for high school students was developed that was composed of six steps, 16 categories, and 36 sub-divisions. The developed program models for each step are as follows: 1)a step involving understanding the decision making process and establishing a strategy, 2)selecting a list of jobs and capabilities (value) through an understanding of oneself, 3)identifying related majors and jobs to explore the major selection, 4)a comparison of majors through the major exploration, 5)the major selection, 6)evaluation and implementation. The finalized college major selection program of adaptability in the field was tested. To test the content validity of the college major selection program, the validity was tested with teachers who had many years of experience in career counseling. 이 연구는 대학전공선택 프로그램 모형 및 각 단계별 내용요소를 구안하고, 구안된 프로그램 모형 및 내용요소의 타당성 검증을 통해 고등학생의 대학전공선택 프로그램 모형을 개발하는 데 목적을 두고 수행되었다. 이러한 연구 목적을 달성하기 위하여 먼저 선행연구에서 제시된 개념을 기초로 고등학생의 대학전공선택의 개념을 도출하였으며, 대학 1학년 학생을 대상으로 실시하는 각 대학의 신입생 실태 및 특성 조사 결과를 분석하여 고등학생의 대학전공선택 특성을 확인하였다. 또한, 대학전공선택 프로그램 모형 및 내용요소를 구안하기 위해 모형 개발 절차, 대학전공선택 관련 변인, 고등학생의 진로발달 과업, 고등학생의 진로발달 능력, 대학전공선택 프로그램 등을 분석하여 6개 단계, 16개 중영역, 36개의 세부영역으로 구성된 프로그램 모형 및 내용요소 초안을 구안하였다. 구안된 프로그램 모형 및 내용요소 초안에 대한 타당성을 검증하기 위하여 2차에 걸친 델파이 조사를 실시하였다. 델파이 조사를 위한 전문가 집단은 교수 및 연구원, 대학진학과 관련된 프로그램 운영 및 업무 담당자, 진학 상담 교사를 각각 5명씩 모두 15명을 전문가 패널로 구성하였다. 매 라운드별 평균, 긍정율, 중위수, 사분범위, CVR값, 합의도, 수렴도, ICC(1), ICC(2) 등을 산출하여 계속 진행에 나갈지를 판단하면서 수행하였다. 그 결과 고등학생의 대학전공선택 프로그램 모형은 1.의사결정 이해 및 전략수립, 2.자신에 대한 이해를 통한 직업목록 및 능력(가치)선정, 3.전공선택을 위한 직업탐색 및 관련 전공파악, 4.전공탐색을 통한 전공비교, 5.전공선택, 6.평가 및 실행의 총 6개 단계, 16개 중영역, 43개 세부영역으로 구성하였다. 개발된 대학전공선택 프로그램 모형의 현장적용가능성 검증을 진학상담 경험이 많은 현직교사를 대상으로 일반화가능도 계수를 활용하여 높은 응답자간의 의견일치지수를 보였다.

Appropriate model selection methods for nonstationary generalized extreme value models

Kim, Hanbeen,Kim, Sooyoung,Shin, Hongjoon,Heo, Jun-Haeng Elsevier 2017 Journal of hydrology Vol.547 No.-

<P><B>Abstract</B></P> <P>Several evidences of hydrologic data series being nonstationary in nature have been found to date. This has resulted in the conduct of many studies in the area of nonstationary frequency analysis. Nonstationary probability distribution models involve parameters that vary over time. Therefore, it is not a straightforward process to apply conventional goodness-of-fit tests to the selection of an appropriate nonstationary probability distribution model. Tests that are generally recommended for such a selection include the Akaike’s information criterion (AIC), corrected Akaike’s information criterion (AICc), Bayesian information criterion (BIC), and likelihood ratio test (LRT). In this study, the Monte Carlo simulation was performed to compare the performances of these four tests, with regard to nonstationary as well as stationary generalized extreme value (GEV) distributions. Proper model selection ratios and sample sizes were taken into account to evaluate the performances of all the four tests. The BIC demonstrated the best performance with regard to stationary GEV models. In case of nonstationary GEV models, the AIC proved to be better than the other three methods, when relatively small sample sizes were considered. With larger sample sizes, the AIC, BIC, and LRT presented the best performances for GEV models which have nonstationary location and/or scale parameters, respectively. Simulation results were then evaluated by applying all four tests to annual maximum rainfall data of selected sites, as observed by the Korea Meteorological Administration.</P> <P><B>Highlights</B></P> <P> <UL> <LI> We compared the AIC, AICc, BIC, and LRT for nonstationary GEV models. </LI> <LI> Monte Carlo simulation was conducted for evaluating the performances of all tests. </LI> <LI> Under stationary conditions, the BIC shows the best performance (N>40). </LI> <LI> Under nonstationary conditions, regression lines for model selection were proposed. </LI> <LI> The results of simulations were verified through the application of observed data. </LI> </UL> </P>

논문 : 융복합 디자인학과의 선발인재상 도출방법에 대한 연구 -UX 특성화 프로그램의 선발인재상 도출 사례를 중심으로-

김정희 ( Jhong Hee Kim ),오동근 ( Dong Keun Oh ),김영준 ( Young Jun Kim ) 디자인융복합학회 2014 디자인융복합연구 Vol.13 No.3

본 연구는 특성화를 목표로 한 디자인학과에서 선발타당도가 높은 선발을 위하여 선발역량들을 도출하고 이들로 구성된 선발인재상이 지원자들의 퍼포먼스를 예측하는 타당한 예측변인임을 검증하고자한다. 이를 통해서 특성화된 디자인 학과에서 학생을 선발하는데 기반이 되는 선발인재상 도출방법을 제안하고자 한다. 이를 위해 디자인 프로세스에 기반하여 적용한 KSAO Competency Modeling Process for Design Student Selection의 단계를 활용하여 학업수행 및 과제수행에서의 성과에 중요한 영향을 미치는 주요행동케이스들을 도출하였다. 이 중에서 성공적 행동사례들에 대한 KSAO분석을 통해 K(knowledge), S(skill), A(ability), O(others)를 도출하였으며, 이들 중 교육의 대상과 선발의 대상을 구분하여 선발역량들로 구성된 선발인재상을 도출하였다. 선발인재상이 신뢰롭고 타당한지를 검증하기 위하여 30명의 졸업생과 재학생들을 대상으로 설문조사를 실시하였으며, t-test를 통해 우수 집단과 비우수 집단 간 인재상에서의 차이를 검증하였으며, 다중회귀분석(multiple regression analysis)을 통해 이들 인재상이 작업에 대한 열정(work passion), 적응도(work adaption), 몰입도(work involvement), 성취도(work achievement)에 미치는 영향을 검증하였다. 또한 이러한 방법론의 활용은 학과의 본래 미션과, 현재의 목표, 미래 산업의 방향 등을 분석하고 예측한다는 관점에서 미래지향적인 융합적 학문연구를 목표로 한 학과의 설립, 구체화된 전문성을 목표로 새로운 도약을 준비하는 대학, 구조적인 선발도구를 설계하여 교육목표에 적합한 인재를 선발하고자 하는 대학에서 활용 가능한 방법으로써 타당도 높은 결과를 이끌어 내도록 도움이 될 것이라 예상한다. This research, in order to select talent with potential for future designers, investigates and verifies the component focused competency model and indicators, and through this, aims to suggest competency model and behavioral indicators to select students for specialized universities. For this study applies job modeling methodology which is a combination of job analysis and competency modeling methodologies, and uses Behavior Indicator Modeling Process to derive at critical incidents that impact academic performance job execution. And after KASO analysis on these successful cases, derive at K(knowledge), S (skill), A (ability), O (others) to classify educational and selecting targets, and arrive at a competency model for student selection and behavioral indicator. After that , in order to verify that the competency model and behavioral indicator are trustworthy, a survey based on thirty graduates and enrolled students were conducted. This research, through KSAO Competency modeling Process methodology, has suggested specific competency model and indicators (1. Perfectionist for Outstanding Craftsmanship, 2. Self-Motivator for Continuous Learning, 3. Knowledge-Integrator for Progressive Performance, 4. Structured Thinker for Delivering Unique Value, 5. Esthete for Visual Harmony) that can be applied during selecting process to screen for students with potential to grow as experts at specialized design universities. From the angle that the university’s past, present, future is analyzed, and predicted, this is an applicable method and also will result in objective and scientific results for a university that aims for a establishment of a future knowledge integrated department, to grow with a specialized objective, and select competent students based on systematic selection tools, and educational objective.

사교육비 지출 패턴과 경감정책의 효과분석 : Tobit Model과 Heckman Selection Model의 활용

이종구,김태진,권기헌 한국교육개발원 2009 한국교육 Vol.36 No.2

The purpose of this study is to explain the factors affecting the expenditure of private tutoring, to analyze the effect of its reduction policy without bias and to suggest the policy implications. For analysis, we put to practical use the Tobit model and the Heckman selection model (2-step), considering the characteristic of private tutoring expenditure as a dependent variable. The analysis showed that private tutoring expenditure was influenced by demographic and socioeconomic characteristics. In particular, we found important implications from factors like household earning, parents' level of scholarship and administrative district. In addition, we deduced that the reduction policy could not result in a visible effect. The 'after-school classes' program had a negative effect on private tutoring expenditure, but this effect was insignificant. Considering the temporal and spatial limitations of the program, we could not say that it had an essential effect on the reduction of private tutoring expenditure. In contrast, the 'internet·broadcast education' program had a positive influence on private tutoring expenditure. The results showed that the 'internet·broadcast education' program acted as a substitute rather than a supplement. Henceforth, for designing the reduction policy of private tutoring expenditure, we should consider the various factors affecting the expenditure of private tutoring and the differential acceptability of the policy according to each of the affecting factors. 본 연구의 목적은 사교육비 지출에 영향을 미치는 요인과 사교육비 경감정책의 효과를 편의 없이 추정하여 정책적 함의를 끌어내는 것이다. 이를 위해 종속변수인 사교육비의 특성을 고려한 Tobit model과 Heckman selection model(2-step)을 분석에 활용하였다. 분석 결과에 따르면 사교육비 지출은 인구학적, 사회경제학적 특성에 따라 많은 차이가 발생하고 있다. 특히 가구의 소득, 부모의 학력 그리고 행정구역등의 요인은 사교육비의 지출에 많은 영향을 미치고 있어 향후 정책 설계에 중요한 시사점을 주고 있다. 한편, 사교육비 경감정책은 가시적인 효과를 보여주지 못하는 것으로 추정된다. 우선 '방과 후 학교' 프로그램은 사교육비 지출과 음의 관계를 갖지만 그 영향력의 크기가 미미하다. 해당 프로그램이 갖는 공간적·시간적 제약을 감안하면 사교육비 경감에 있어 본질적인 효과가 있는 것은 아니다. 다음으로 '인터넷·방송 교육’의 경우 사교육비 지출과 양의 관계를 보이고 있어 사교육에 대해 대체재가 아닌 보완재로 기능하는 것으로 추측된다. 향후의 사교육비 경감정책은 대상 집단의 사교육비 지출에 영향을 미치는 다양한 요인과 그에 따른 정책수용성의 차이를 고려해야 할 것이다.

광고 모델 관련 광고 노이즈가 브랜드 인지도와 브랜드 태도에 미치는 영향

정재학(Jaihak Chung),이상미(Sang-Mi Lee) 한국마케팅과학회 2008 마케팅과학연구 Vol.18 No.3

  광고를 제작, 집행하는 과정에서 기업은 때로 본의 아니게 소비자에게 흔동을 야기 시켜 해당 제품이 무엇인지 잘 못 인지시키거나 제품 이미지를 잘 못 이해하게 하는 경우가 있다. 본 연구는 광고를 제작, 집행하는 과정에서 소비자가 특정 광고를 인지하고 이해하는 데 혼동을 일으키는 모든 제반 요소를 노이즈라고 정의하고 이 요인이 실재 존재하는 지, 또 광고 효과에 일마나 영향을 미치는 지를 알아보고자 한다. 본 연구에서는 가능한 여러 형태의 광고 노이즈 현상 중 특히 한 모델이 동일 시점에서 여러 제품의 광고에 중복 출연함으로 인해 특정 광고에 대한 소비자 혼란을 야기시키는 광고 중복 출연에 따른 노이즈 현상과 한 제품의 광고에서 기용한 현재 광고모델이 과거 광고 모델과 다른 이미지를 가지고 있음으로 인해 해당 광고에 대한 소비자 혼란을 야기 시키는 모델 교체에 따른 광고 노이즈 현상을 연구 하고자 한다. 더 나아가 산업에서 많이 나타나고 있는 모델의 광고 중복 출연과 동일 제품에 대한 광고 모델을 교체하는 것이 소비자 혼란을 일으킬 수 있고 소비자 혼란은 결국, 광고 효과에 부정적인 영항을 미칠 수 있다. 본 연구는 이와 더불어 여섯 가지 조절 변수를 찾아내어 광고 노이즈가 광고 효과에 미치는 부정적인 영항이 어느 경우에 더욱 커지거나 또는 감소하는 지를 알아보고자 하였다.<BR>  광고 중복 출연에 따른 노이즈가 광고 효과에 미치는 영항을 실증 분석한 결과를 해석하면 다음과 같이 정리할 수 있다. 첫째, 동시에 여러 광고에 출연하는 모델을 자사 브랜드광고에 기용하는 것은 브랜드 인지도와 브랜드 태도를 향상하는 데 모두 부정적인 역할을 한다. 둘째, 하지만, 광고 중복도가 높은 모델을 이용해야 한다면 특히 브랜드 인지도에 미치는 부정적인 영항을 줄이기 위해서 제품 이미지와 모델 이미지와의 적합성 정도가 높은 광고모델을 선정하는 것이 바람직하다. 셋째, 광고 중복도가 높은 모델을 이용할 경우, 자사 제품 광고에서 해당 모델이 유지해야 할 이미지가 해당 모델이 중복 출연한 광고 속에서 유지하는 이미지와 달리 하는 것이 자사 제품 인지도 향상에 도움이 되나 중요한 것은 자사 제품 태도 향상에는 효과 차이가 없다는 점이다. 넷째, 자사 모델이 중복 출연한 광고의 제품과 자사 제품이 유사할수록 모델의 광고 중복도는 브랜드 인지도와 브랜드 태도에 모두 부정적 영항을 미친다.<BR>  또한, 모델 교체로 인한 소비자 혼란이 광고 효과에 미치는 영향을 실증 분석한 결과를 해석하면 다음과 같다 첫째, 한 제품의 광고에 여러 모델을 기용할수록 브랜드 인지도에 부정적 영항을 미치지만 브랜드 태도에는 오히려 긍정적 영항을 미치고 있었다. 둘째, 기존 광고 모델보다 현재 광고모델의 광고 적합성이 높으면 제품 광고의 모델 중복도가 브랜드에 미치는 부정적 영향은 약화되지만 브랜드 태도에 모델 중복도가 미치는 영항을 조절하지는 못했다. 셋째, 기존 모델과 현재 모델과의 이미지기 유사하면 모델 중복이 브랜드 인지도에 부정적인 영항을 미치지만 브랜드 태도에는 긍정적 영항을 미친다. 마지막으로 기존 광고와 현재 광고의 컨셉이 유사할 수록 제품 광고의 모델 중복도가 브랜드 인지도에 미치는 영항은 긍정적이었다. 특히, 본 연구에서 살펴본 두 가지 광고 노이즈 현상, 즉 동일 모델의 광고 중복 출연이 모델 교체보다 광고 효과에 부정적인 영항을 미치는 경향이 있으며, 광고 노이즈 현상은 브랜드 태도보다는 브랜드 인지도 형성에 더 뚜렷한 영항을 미침을 알 수 있었다.   Most of the extant studies on communication effects have been devoted to the typical issue, "what types of communication activities are more effective for brand awareness or brand attitudes?" However, little research has addressed another question on communication decisions, "what makes communication activities less effective?"<BR>  Our study focuses on factors negatively influenced on the efficiency of communication activities, especially of Advertising. Some studies have introduced concepts closely related to our topic such as consumer confusion, brand confusion, or belief confusion. Studies on product belief confusion have found some factors misleading consumers to misunderstand the physical features of products. Studies on brand confusion have uncovered factors making consumers confused on brand names. Studies on advertising confusion have tested the effects of ad models" employed by many other firms for different products on communication efficiency.<BR>  We address a new concept, Ad noises, which are any factors interfering with consumers exposed to a particular advertisement in understanding messages provided by advertisements. The objective of this study is to understand the effects of ad noises caused by ad models on brand awareness and brand attitude.<BR>  There are many different types of AD noises. Particularly, we study the effects of AD noises generated from ad model selection decision. Many companies want to employ celebrities as AD models while the number of celebrities who command a high degree of public and media attention are limited. Inevitably, several firms have been adopting the same celebrities as their AD models for different products. If the same AD model is adopted for TV commercials for different products, consumers exposed to those TV commercials are likely to fail to be aware of the target brand due to interference of TV commercials, for other products, employing the same AD model. This is an ad noise caused by employing ad models who have been exposed to consumers in other advertisements, which is the first type of ad noises studied in this research.<BR>  Another type of AD noises is related to the decision of AD model replacement for the same product advertising. Firms sometimes launch another TV commercial for the same products. Some firms employ the same AD model for the new TV commercial for the same product and other firms employ new AD models for the new TV commercials for the same product. The typical problem with the replacement of AD models is the possibility of interfering with consumers in understanding messages of the TV commercial due to the dissimilarity of the old and new AD models.<BR>  We studied the effects of these two types of ad noises, which are the typical factors influencing on the effect of communication : (1) ad noises caused by employing ad models who have been exposed to consumers in other advertisements and (2) ad noises caused by changing ad models with different images for same products. First, we measure the negative influence of AD noises on brand awareness and attitudes, in order to provide the importance of studying AD noises.<BR>  Furthermore, our study unveiled the mediating conditions(variables) which can increase or decrease the effects of ad noises on brand awareness and attitudes.<BR>  We study the effects of three mediating variables for ad noises caused by employing ad models who have been exposed to consumers in other advertisements: (1) the fit between product image and AD model image, (2) similarity between AD model images in multiple TV commercials employing the same AD model, and (3) similarity between products of which TV commercial employed the same AD model. We analyze the effects of another three mediating variables for ad noises caused by changing ad models with different images for same products: (l) the fit of old and new AD models for the same pr

발명영재 학생 선발모형 연구

최유현,심재영,이한규 한국실과교육학회 2008 한국실과교육학회지 Vol.21 No.1

The purposes of the study were to reach the educational specialists' agreement on selection for gifted students in invention, to draw a selection model, and to suggest a practical strategy. We carried out a survey on selection steps and detail process to 32 educational specialists'. After that, we held professional meeting twice and a seminar to draw a selection model. The results of this study were as follows.First, The survey of educational specialists' showed four steps selection process for gifted students in invention. We should consider student's interest and aptitude in first step. Then, the cognitive ability and personality were important factors to screen the gifted. In third step, we should observe student's specificity on invention. Lastly, we had to evaluate student's portfolio.Second, we proposed four steps' selection model in accordance with the result of survey.Third, we suggested a practical strategy on selection for gifted students in invention. The strategy included a study on students' characteristics and selection steps, an experimental application, and an analysis on selection criteria and process. 이 연구의 목적은 발명영재 선발의 방향과 선발방안에 대한 추진전략을 제시하는 것이다.연구의 방법은 설문지를 통해 영재교육전문가 32명에게 발명영재 선발 단계와 세부과정에 대하여 조사하였고, 두 번에 걸친 전문가 협의회와 세미나를 개최하여 내용을 수렴하였다. 연구의 결과는 다음과 같다.첫째, 발명영재 학생 선발 방안에 대한 전문가들의 의견 수렴을 위한 설문조사 결과, 1단계에서는 발명에 대한 학생의 흥미와 적성을 중요하게 생각하고 있었다. 2단계는 주로 학생의 인지적 사고능력 및 성향검사가 필요하다고 하였다. 3단계는 행동관찰, 면담, 집단 토론, 실기 등 실습을 통한 전문성을 평가할 필요가 있다는 의견으로 보였다. 4단계에서는 교육 프로그램을 통한 수행관찰, 산출물 평가, 동료 평가, 수업태도 평가, 개인적 성향, 만족도 평가 등을 고려하고 있음을 알 수 있었다. 둘째, 발명영재 학생의 선발모형 도출 결과는 4단계로 구성하였다. 셋째, 발명영재 학생의 선발을 위한 추진전략을 제안하였다. 발명영재의 선발은 처음 실시되는 만큼 선발대상학생의 특성과 선발 단계 및 방법에 대한 연구가 우선적으로 필요하다. 또한 개발된 선발도구에 대한 실험적 적용을 통하여, 선발도구를 수정 및 보완하는 과정이 필요할 것이다.

Extending the Scope of Automatic Time Series Model Selection: The Package autots for R

Jang, Dong-Ik,Oh, Hee-Seok,Kim, Dong-Hoh The Korean Statistical Society 2011 Communications for statistical applications and me Vol.18 No.3

In this paper, we propose automatic procedures for the model selection of various univariate time series data. Automatic model selection is important, especially in data mining with large number of time series, for example, the number (in thousands) of signals accessing a web server during a specific time period. Several methods have been proposed for automatic model selection of time series. However, most existing methods focus on linear time series models such as exponential smoothing and autoregressive integrated moving average(ARIMA) models. The key feature that distinguishes the proposed procedures from previous approaches is that the former can be used for both linear time series models and nonlinear time series models such as threshold autoregressive(TAR) models and autoregressive moving average-generalized autoregressive conditional heteroscedasticity(ARMA-GARCH) models. The proposed methods select a model from among the various models in the prediction error sense. We also provide an R package autots that implements the proposed automatic model selection procedures. In this paper, we illustrate these algorithms with the artificial and real data, and describe the implementation of the autots package for R.