• Journal of the Korean Data and Information Science Society
• /
• 제27권1호
• /
• pp.111-120
• /
• 2016

대부분의 불연속 회귀함수의 커널추정량은 알고 있거나 추정된 불연속점을 기준으로 자료를 분리하여 각각을 독립적으로 회귀함수를 적합하고 있다. 회귀모형에서 분산함수가 불연속점을 가지고 있을 때에도 잔차제곱들을 이용하여 위와 같은 불연속 회귀함수의 커널추정법을 활용하고 있다. Kang 등 (2000)은 $M{\ddot{u}}ller$ (1992)의 불연속점과 점프크기 커널추정량을 이용하여 반응변수의 표본을 연속인 회귀함수로부터 표본인 것처럼 수정하여 불연속 회귀함수를 추정하였다. 본 연구에서는 불연속 분산함수를 추정하기 위하여 Kang 등 (2000)의 방법을 이용한다. Kang과 Huh (2006)의 분산함수의 불연속점과 점프크기 추정량으로 잔차제곱들을 수정하고, 수정된 잔차제곱들을 이용하여 불연속 분산함수 커널추정량을 제안할 것이다. 제안된 추정량의 적분제곱오차의 수렴속도를 보여주고 모의실험을 통하여 기존의 추정량과 제안된 추정량을 비교하고자 한다.

• 응용통계연구
• /
• 제37권1호
• /
• pp.39-48
• /
• 2024

이 논문에서는 비대칭 지렛대 효과를 처리하기 위한 확장된 GARCH 모형을 소개한다. 표준 GARCH 모형의 분산은 이전의 조건부 분산과 이전의 잔차 제곱 항으로 구성되어 있다. 잔차 제곱항으로는 비대칭 지렛대 효과를 모형화할 수 없는데 이 논문에서는 Yeo-Johnson 변환을 이용하여 지렛대 효과를 설명하는 확장된 GARCH 모형을 제안한다. Yeo-Johnson 변환은 변환 모수를 조절하여 비대칭 자료를 보다 정규성 또는 대칭성을 만족하도록 만든데 우리는 Yeo-Johnson 변환의 성질을 역으로 이용하여 비대칭 변동성을 모형화 한다. 제안 모형의 특징에 대해 살펴보고 모수 추정에 대해 알아본다. 제안 모형에서 예측과 예측구간을 어떻게 구하는지 살펴보고 실증 자료분석을 통해 제안모형과 GARCH, 비대칭 지렛대 효과를 모형화한 다른 GARCH 모형을 비교해 본다.

• Journal of Biosystems Engineering
• /
• 제43권4호
• /
• pp.369-378
• /
• 2018

Purpose: The aim of this study was to construct a saponin content-predicting model using shortwave infrared imaging spectroscopy. Methods: The experiment used a shortwave imaging spectrometer and ENVI spectral acquisition software sampling a spectrum of 910 nm-2500 nm. The corresponding preprocessing and mathematical modeling analysis was performed by Unscrambler 9.7 software to establish a ginsenoside nondestructive spectral testing prediction model. Results: The optimal preprocessing method was determined to be a standard normal variable transformation combined with the second-order differential method. The coefficient of determination, $R^2$, of the mathematical model established by the partial least squares method was found to be 0.9999, while the root mean squared error of prediction, RMSEP, was found to be 0.0043, and root mean squared error of calibration, RMSEC, was 0.0041. The residuals of the majority of the samples used for the prediction were between ${\pm}1$. Conclusion: The experiment showed that the predicted model featured a high correlation with real values and a good prediction result, such that this technique can be appropriately applied for the nondestructive testing of ginseng quality.

• The Journal of Asian Finance, Economics and Business
• /
• 제9권7호
• /
• pp.1-15
• /
• 2022

Combining the strength of both Mixed Data Sampling (MIDAS) Regression and realized variance measures, this paper seeks to investigate two objectives: (1) evaluate the post-sample performance of the proposed weekly Realized Variance-MIDAS (RVar-MIDAS) in one-week ahead volatility forecasting against the established Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model and the less explored but robust STES (Smooth Transition Exponential Smoothing) methods. (2) comparing forecast error performance between realized variance and squared residuals measures as a proxy for actual volatility. Data of seven private equity mutual fund indices (generated from 57 individual funds) from two different time periods (with and without financial crisis) are applied to 21 models. Robustness of the post-sample volatility forecasting of all models is validated by the Model Confidence Set (MCS) Procedures and revealed: (1) The weekly RVar-MIDAS model emerged as the best model, outperformed the robust DAILY-STES methods, and the weekly DAILY-GARCH models, particularly during a volatile period. (2) models with realized variance measured in estimation and as a proxy for actual volatility outperformed those using squared residual. This study contributes an empirical approach to one-week ahead volatility forecasting of mutual funds return, which is less explored in past literature on financial volatility forecasting compared to stocks volatility.

• 응용통계연구
• /
• 제24권5호
• /
• pp.759-768
• /
• 2011

Tree algorithms have been widely developed for regression problems. One of the good features of a regression tree is the flexibility of fitting because it can correctly capture the nonlinearity of data well. Especially, data with sudden structural breaks such as the price of oil and exchange rates could be fitted well with a simple mixture of a few piecewise linear regression models. Now that split points are determined by chi-squared statistics related with residuals from fitting piecewise linear models and the split variable is chosen by an objective criterion, we can get a quite reasonable fitting result which goes in line with the visual interpretation of data. The piecewise linear regression by a regression tree can be used as a good fitting method, and can be applied to a dataset with much fluctuation.

• Journal of the Korean Data and Information Science Society
• /
• 제20권5호
• /
• pp.925-931
• /
• 2009

In this paper we propose a variance function estimation method for replicated data based on averages of squared residuals obtained from estimated mean function by the least squares support vector machine. Newton-Raphson method is used to obtain associated parameter vector for the variance function estimation. Furthermore, the cross validation functions are introduced to select the hyper-parameters which affect the performance of the proposed estimation method. Experimental results are then presented which illustrate the performance of the proposed procedure.

• Communications for Statistical Applications and Methods
• /
• 제20권1호
• /
• pp.41-52
• /
• 2013

The stationary bootstrap of Politis and Romano (1994) is adopted to develop prediction intervals of returns and volatilities in a generalized autoregressive heteroskedastic (GARCH)(p, q) model. The stationary bootstrap method is applied to generate bootstrap observations of squared returns and residuals, through an ARMA representation of the GARCH model. The stationary bootstrap estimators of unknown parameters are defined and used to calculate the stationary bootstrap samples of volatilities. Estimates of future values of returns and volatilities in the GARCH process and the bootstrap prediction intervals are constructed based on the stationary bootstrap; in addition, asymptotic validities are also shown.

• Coupled systems mechanics
• /
• 제6권1호
• /
• pp.55-74
• /
• 2017

The evaluation of torsional effects on multistory buildings remains an open issue, despite considerable research efforts and numerous publications. In this study, a large number of multiple test structures are considered with normally distributed topological attributes, in order to quantify the statistically derived relationships between the torsional criteria and response parameters. The linear regression analysis results, depict that the center of twist and the ratio of torsion (ROT) index proved numerically to be the most reliable criteria for the prediction of the modal rotation and displacements, however the residuals distribution and R-squared derived for the ductility demands prediction, was not constant and low respectively. Thus, the assessment of the torsional parameters' contribution to the nonlinear structural response was investigated using artificial neural networks. Utilizing the connection weights approach, the Center of Strength, Torsional Stiffness and the Base Shear Torque curves were found to exhibit the highest impact numerically, while all the other torsional indices' contribution was investigated and quantified.

• Communications for Statistical Applications and Methods
• /
• 제28권5호
• /
• pp.537-551
• /
• 2021

We conducted a study on a regression spline estimator with a few pre-specified auxiliary variables. For the implementation of the proposed estimators, we adapted a coordinate descent algorithm. This was implemented by considering a structure of the sum of the residuals squared objective function determined by the B-spline and the auxiliary coefficients. We also considered an efficient stepwise knot selection algorithm based on the Bayesian information criterion. This was to adaptively select smoothly functioning estimator data. Numerical studies using both simulated and real data sets were conducted to illustrate the proposed method's performance. An R software package psav is available.

• 재무관리연구
• /
• 제11권1호
• /
• pp.233-262
• /
• 1994

본 연구는 주식시장(株式市場)의 이상현상(異狀現象)중의 하나인 요일효과(曜日效果)(day of the week effect)를 전통적인 회귀분석(回歸分析)이 아닌 ARCH 또는 GARCH 모형을 사용하여 조건부(條件附) 평균수익률(기대수익률)(平均收益率(期待收益率)) 뿐만아니라 조건부(條件附) 분산(分散)에도 나타나는지에 대하여 분석하였으며, 규모별(規模別)에 따라 요일효과(曜日效果)에 어떠한 차이가 나타나는지를 분석하였다. 본 연구의 추정결과를 요약하면, 조건부(條件附) 평균수익률(기대수익률)(平均收益率(期待收益率)) 및 조건부(條件附) 분산(分散) 모두에 있어 요일효과(曜日效果)가 뚜렷하게 존재하는 것으로 나타났다. 즉, 조건부(條件附) 평균수익률(平均收益率)에 대해서는 월요일(月曜日)은 부(負)의 효과, 토요일(土曜日)은 정(正)의 효과가 나타났으며, 조건부(條件附) 분산(分散)에 대해서는 월요일(月曜日)은 정(正)의 효과가, 토요일(土曜日)은 부(負)의 효과가 발견되었다. 그러나 한국(韓國)의 주식시장의 본격적인 성장기이면서 주식가격의 등락이 심했던 $86\sim92$년(年)간의 표본기간 동안에는 조건부(條件附) 분산(分散)에 대한 요일효과(曜日效果)는 존재하였으나, 조건부(條件附) 평균수익률(平均收益率)에 대한 요일효과(曜日效果)는 존재하지 않는 것으로 나타났다. 그리고 소형지수(小型指數)가 중(中) 대형지수(大型指數)와는 다른 주가행태를 보이는 것으로 나타났으며, 다음과 같은 몇 가지의 규모별(規模別) 차이(差異)를 보였다. 첫째, 조건부(條件附) 평균수익률(平均收益率)에 대한 분석에서 중(中) 대형지수수익률(大型指數收益率)을 사용하였을 경우에는 요일효과(曜日效果)가 나타난 반면에, 소형(小型) 지수수익률(指數收益率)의 경우에는 화요효과(火曜效果)가 존재하는 것으로 나타났다. 둘째, 조건부(條件附) 분산(分散)에 대한 분석에서 정(正)의 공휴일효과(公休日效果)가 다른 규모별 지수수익률(指數收益率)의 경우에는 나타나지 많았지만 소형(小型) 지수수익률(指數收益率)의 경우에는 존재하는 것으로 나타났다. 세째, 소형(小型) 지수수익률(指數收益率)의 경우 모형 추정후의 정규잔차(定規殘差)(normalized residuals) 및 정규자승잔차(定規自乘殘差)(normalized squared residuals)에 대한 시계열상관(時系列相關) 검정결과 모형의 부적합성(不適合性)이 나타났다. 본 연구는 기존의 기대수익률(期待收益率) 위주의 요일효과(曜日效果) 분석에서 주식수익률(株式收益率)의 분산(分散) 즉, 변동성(變動性)에 촛점을 두어 분석하였으며, 이는 투자자의 정확한 위험측정(危險測定)수단의 제공이라는 면에서 의의(意義)가 있을 것으로 생각된다.