• Title/Summary/Keyword: quasi-likelihood estimation

Search Result 18, Processing Time 0.024 seconds

THRESHOLD MODELING FOR BIFURCATING AUTOREGRESSION AND LARGE SAMPLE ESTIMATION

  • Hwang, S.Y.;Lee, Sung-Duck
    • Journal of the Korean Statistical Society
    • /
    • v.35 no.4
    • /
    • pp.409-417
    • /
    • 2006
  • This article is concerned with threshold modeling of the bifurcating autoregressive model (BAR) originally suggested by Cowan and Staudte (1986) for tree structured data of cell lineage study where each individual $(X_t)$ gives rise to two off-spring $(X_{2t},\;X_{2t+1})$ in the next generation. The triplet $(X_t,\;X_{2t},\;X_{2t+1})$ refers to mother-daughter relationship. In this paper we propose a threshold model incorporating the difference of 'fertility' of the mother for the first and second off-springs, and thereby extending BAR to threshold-BAR (TBAR, for short). We derive a sufficient condition of stationarity for the suggested TBAR model. Also various inferential methods such as least squares (LS), maximum likelihood (ML) and quasi-likelihood (QL) methods are discussed and relevant limiting distributions are obtained.

Extended Quasi-likelihood Estimation in Overdispersed Models

  • Kim, Choong-Rak;Lee, Kee-Won;Chung, Youn-Shik;Park, Kook-Lyeol
    • Journal of the Korean Statistical Society
    • /
    • v.21 no.2
    • /
    • pp.187-200
    • /
    • 1992
  • Samples are often found to be too heterogeneous to be explained by a one-parameter family of models in the sense that the implicit mean-variance relationship in such a family is violated by the data. This phenomenon is often called over-dispersion. The most frequently used method in dealing with over-dispersion is to mix a one-parameter family creating a two parameter marginal mixture family for the data. In this paper, we investigate performance of estimators such as maximum likelihood estimator, method of moment estimator, and maximum quasi-likelihood estimator in negative binomial and beta-binomial distribution. Simulations are done for various mean parameter and dispersion parameter in both distributions, and we conclude that the moment estimators are very superior in the sense of bias and asymptotic relative efficiency.

  • PDF

Estimation of Spatial Dependence with GEE

  • Lee, Yoon-Dong;Choi, Hye-Mi
    • Proceedings of the Korean Statistical Society Conference
    • /
    • 2003.05a
    • /
    • pp.269-273
    • /
    • 2003
  • We consider an efficient parametric estimation method of spatial dependence in weak stationary processes. Spatial dependence is modeled through variogram and correlogram. Most of parametric estimation methods of correlogram use two step method; nonparametric estimation and parametric integration. We bind these two steps into one step by using GEE method instead of least squares type optimization. Our one step method is more efficient statistically and gives a clear interpretation of related concepts used in traditional two step methods.

  • PDF

Sequential Adaptation Algorithm Based on Transformation Space Model for Speech Recognition (음성인식을 위한 변환 공간 모델에 근거한 순차 적응기법)

  • Kim, Dong-Kook;Chang, Joo-Hyuk;Kim, Nam-Soo
    • Speech Sciences
    • /
    • v.11 no.4
    • /
    • pp.75-88
    • /
    • 2004
  • In this paper, we propose a new approach to sequential linear regression adaptation of continuous density hidden Markov models (CDHMMs) based on transformation space model (TSM). The proposed TSM which characterizes the a priori knowledge of the training speakers associated with maximum likelihood linear regression (MLLR) matrix parameters is effectively described in terms of the latent variable models. The TSM provides various sources of information such as the correlation information, the prior distribution, and the prior knowledge of the regression parameters that are very useful for rapid adaptation. The quasi-Bayes (QB) estimation algorithm is formulated to incrementally update the hyperparameters of the TSM and regression matrices simultaneously. Experimental results showed that the proposed TSM approach is better than that of the conventional quasi-Bayes linear regression (QBLR) algorithm for a small amount of adaptation data.

  • PDF

Improvement of Suspended Solid Loads Estimation in Nakdong River Using Minimum Variance Unbiased Estimator (비편향 회귀분석모형을 이용한 낙동강 본류 부유사량 산정방법의 신뢰도 향상)

  • Han, Suhee;Kang, Du Kee;Shin, Hyun Suk;Yu, Jae-Jeong;Kim, Sangdan
    • Journal of Korean Society on Water Environment
    • /
    • v.23 no.2
    • /
    • pp.251-259
    • /
    • 2007
  • In this study three log-transformed linear regression models are compared with the focus of bias correction problem. The models are the traditional simple linear regression estimator (SL), the quasi maximum likelihood estimator (QMLE) and the minimum variance unbiased estimator (MVUE). Using such models, suspended solid loads can be estimated using the discharge - suspended solid data set that has been measured by NIER Nakdong River Water Environment Laboratory. As a result, SL shows negative bias for most values of the measured discharge range. QMLE is nearly unbiased for moderate values of the measured discharge range, but shows increasingly positive bias for either large or small value of the measured discharge range. MVUE is unbiased. It is also analyzed how the estimated regression coefficient and exponent are distributed along Nakdong river main stream.

Comparison of semiparametric methods to estimate VaR and ES (조건부 Value-at-Risk와 Expected Shortfall 추정을 위한 준모수적 방법들의 비교 연구)

  • Kim, Minjo;Lee, Sangyeol
    • The Korean Journal of Applied Statistics
    • /
    • v.29 no.1
    • /
    • pp.171-180
    • /
    • 2016
  • Basel committee suggests using Value-at-Risk (VaR) and expected shortfall (ES) as a measurement for market risk. Various estimation methods of VaR and ES have been studied in the literature. This paper compares semi-parametric methods, such as conditional autoregressive value at risk (CAViaR) and conditional autoregressive expectile (CARE) methods, and a Gaussian quasi-maximum likelihood estimator (QMLE)-based method through back-testing methods. We use unconditional coverage (UC) and conditional coverage (CC) tests for VaR, and a bootstrap test for ES to check the adequacy. A real data analysis is conducted for S&P 500 index and Hyundai Motor Co. stock price index data sets.

Additive hazards models for interval-censored semi-competing risks data with missing intermediate events (결측되었거나 구간중도절단된 중간사건을 가진 준경쟁적위험 자료에 대한 가산위험모형)

  • Kim, Jayoun;Kim, Jinheum
    • The Korean Journal of Applied Statistics
    • /
    • v.30 no.4
    • /
    • pp.539-553
    • /
    • 2017
  • We propose a multi-state model to analyze semi-competing risks data with interval-censored or missing intermediate events. This model is an extension of the three states of the illness-death model: healthy, disease, and dead. The 'diseased' state can be considered as the intermediate event. Two more states are added into the illness-death model to incorporate the missing events, which are caused by a loss of follow-up before the end of a study. One of them is a state of the lost-to-follow-up (LTF), and the other is an unobservable state that represents an intermediate event experienced after the occurrence of LTF. Given covariates, we employ the Lin and Ying additive hazards model with log-normal frailty and construct a conditional likelihood to estimate transition intensities between states in the multi-state model. A marginalization of the full likelihood is completed using adaptive importance sampling, and the optimal solution of the regression parameters is achieved through an iterative quasi-Newton algorithm. Simulation studies are performed to investigate the finite-sample performance of the proposed estimation method in terms of empirical coverage probability of true regression parameters. Our proposed method is also illustrated with a dataset adapted from Helmer et al. (2001).

Relation between Risk and Return in the Korean Stock Market and Foreign Exchange Market (주가와 환율의 위험-수익 관계에 대한 연구)

  • Park, Jae-Gon;Lee, Phil-Sang
    • The Korean Journal of Financial Management
    • /
    • v.26 no.3
    • /
    • pp.199-226
    • /
    • 2009
  • We examine the intertemporal relation between risk and return in the Korean stock market and foreign exchange market based on the two factor ICAPM framework. The standard GARCH model and the GJR(1993) model are employed to estimate conditional variances of the stock returns and foreign exchange rates. The covariance between the rates of stock returns and changes in the exchange rates are estimated by the constant conditional correlation model of Bollerslev(1990) and the dynamic conditional correlation model of Engle(2002). The multivariate GARCH in mean model and quasi-maximum likelihood estimation method, consequently, are applied to investigate riskreturn relation jointly. We find that the estimated coefficient of relative risk aversion is negative and statistically significant in the post-financial crisis sample period in the Korean stock market. We also show that the expected stock returns are negatively related to the dynamic covariance with foreign exchange rates. Both estimated parameters of conditional variance and covariance in the foreign exchange market, however, are not statistically significant. The GJR model is better than the standard GARCH model to estimate the conditional variances. In addition, the dynamic conditional correlation model has higher explanatory power than the constant correlation model. The empirical results of this study suggest following two points to investors and risk managers in hedging and diversifying strategies for their portfolios in the Korean stock market: first, the variability of foreign exchange rates should be considered, and second, time-varying correlation between stock returns and changes in foreign exchange rates supposed to be considered.

  • PDF