| 1. |
- Mathew, Thomas, et al.
(author)
-
Nonnegative estimation of variance components in heteroscedastic one-way random effects ANOVA models
- 2007
-
In: Research Report, Centre of Biostochastics, Swedish University of Agricultural Sciences. - Umeå : Centre of Biostochastics. - 1651-8543. ; :1
-
Reports (other academic/artistic)abstract
- There is a considerable amount of literature dealing with inference about the parameters in a heteroscedastic one-way random-effects ANOVA model. In this paper, we primarily address the problem of improved quadratic estimation of the random-effect variance component. It turns out that such estimators with a smaller mean squared error compared with some standard unbiased quadratic estimators exist under quite general conditions. Improved estimators of the error variance components are also established.
|
|
| 2. |
- Nahtman, Tatjana, et al.
(author)
-
On a Class of Singular Nonsymmetric Matrices with Nonnegative Integer Spectra
- 2007
-
Reports (other academic/artistic)abstract
- The objective of this paper is to consider a class of singular nonsymmetric matrices with integer spectrum. The class comprises generalized triangular matrices with diagonal elements obtained by summing the elements of the corresponding column. If the size of a matrix belonging to the class equals $n\times n$, the spectrum of the matrix is given by the sequence of distinct non-negative integers up to n-1, irrespective of the elements of the matrix. Right and left eigenvectors are obtained. Moreover, several interesting relations are presented, including factorizations via triangular matrices.
|
|
| 3. |
- Srivastava, Muni S., et al.
(author)
-
Estimation in General Multivariate Linear Models with Kronecker Product Covariance Structure
- 2008
-
Reports (other academic/artistic)abstract
- In this article models based on pq-dimensional normally distributed random vectors x are studied with a mean vec(ABC), where A and Care known matrices, and a separable covariance matrix $\psi\otimes \Sigma$, where both $\Psi$ and $\Sigma$ are positive definite and except the estimability condition $\psi_{qq} = 1$, unknown. The model may among others be applied when spatial-temporal relationships exist. On the basis of n independent observations on the random vector x, we wish to estimate the parameters of the model. In the paper estimation equations for obtaining maximum likelihood estimators are presented. It is shown that there exist only one solution to these equations. Likelihood equations are also considered when $FBG = 0$, with F and G known. Moreover, the likelihood ratio test for testing $FBG = 0$ against $FBG\neq = 0$ is considered.
|
|
| 4. |
|
|
| 5. |
|
|
| 6. |
- Cronie, Ottmar, et al.
(author)
-
Maximum likelihood estimation in a discretely observed immigration-death process
- 2010
-
Reports (other academic/artistic)abstract
- In order to find the maximum likelihood (ML) estimator of the parameter pair governing the immigration-death process (a continuous time Markov chain) we derive its transition probabilities. The likelihood maximisation problem is reduced from two dimensions to one dimension. We also show the consistency and the asymptotic normality of the ML-estimator under an equidistant sampling scheme, given that the parameter pair lies in some compact subset of the positive part of the real plane. We thereafter evaluate, numerically, the behaviour of the estimator and we finally see how our ML-estimation can be applied to the so-called Renshaw-Särkkä growth interaction model; a spatio-temporal point process with time dependent interacting marks in which the immigration-death process controls the arrivals of new marked points as well as their potential life-times.
|
|
| 7. |
- Ekström, Magnus
(author)
-
Nonparametric estimation for classic and interval open-ended data in contingent valuation
- 2010
-
Reports (other academic/artistic)abstract
- Assume that a valuation survey is conducted for estimating the benefits of a policy change. In order to estimate the willingness to pay (WTP) for the policy change, a classic and interval open-ended question is used. This question permits respondents to choose between two types of answer, either an exact WTP amount or a self-selected interval. Accidently, a rather large proportion of the respondents give exact WTP values as well as intervals. In this set up, we derive a nonparametric maximum likelihood estimator of the distribution of WTP. Monte Carlo simulations are performed to study the behavior of the proposed estimator
|
|
| 8. |
- Kollo, Tõnu, et al.
(author)
-
Estimation in high-dimensional analysis and multivariate linear models
- 2008
-
Reports (other academic/artistic)abstract
- When the number of variables compared with the number of observationsis large this paper presents a new approach of estimating the parametersdescribing the mean structure in the Growth Curve model. An explicitestimator is obtained which is unbiased, consistent and asymptoticallynormally distributed. Its variance is also derived.
|
|
| 9. |
- Kuljus, Kristi, et al.
(author)
-
Modeling Swedish Electricity Prices for 2000-2009
- 2010
-
Reports (other academic/artistic)abstract
- This work presents an empirical approach to modeling Swedish electricity prices, where the expected mean price will be independent of past prices. Linear regression is used for modeling the mean price and autoregressive time series models for describing the behavior of the regression residuals. The estimated model can be used for investigating what influence the explanatory variables in the regression model have on the Swedish area price
|
|
| 10. |
|
|
