scholarly journals On the solution of Stein’s equation and Fisher’s information matrix of an ARMAX process

2005 ◽  
Vol 396 ◽  
pp. 1-34 ◽  
Author(s):  
André Klein ◽  
Peter Spreij
Keyword(s):  
Information Matrix ◽  
Fisher's Information

Efficient Block Training of Multilayer Perceptrons

2000 ◽  
Vol 12 (6) ◽  
pp. 1429-1447 ◽  
Author(s):  
A. Navia-Vázquez ◽  
A. R. Figueiras-Vidal
Keyword(s):  
Correction Factor ◽  
Sample Selection ◽  
Information Matrix ◽  
Computational Effort ◽  
Second Order ◽  
Multilayer Perceptrons ◽  
Training Algorithms ◽  
Recurrent Networks ◽  
Natural Gradient ◽  
Fisher's Information

The attractive possibility of applying layerwise block training algorithms to multilayer perceptrons MLP, which offers initial advantages in computational effort, is refined in this article by means of introducing a sensitivity correction factor in the formulation. This results in a clear performance advantage, which we verify in several applications. The reasons for this advantage are discussed and related to implicit relations with second-order techniques, natural gradient formulations through Fisher's information matrix, and sample selection. Extensions to recurrent networks and other research lines are suggested at the close of the article.

Optimal Calibration Designs for Tests of Polytomously Scored Items Described by Item Response Theory Models

2001 ◽  
Vol 26 (4) ◽  
pp. 361-380 ◽  
Author(s):  
Rebecca Holman ◽  
Martijn P. F. Berger
Keyword(s):  
Item Response Theory ◽  
Item Response ◽  
Rating Scale ◽  
Information Matrix ◽  
Scale Model ◽  
Partial Credit Model ◽  
Response Theory ◽  
Irt Models ◽  
Optimal Calibration ◽  
Fisher's Information

This article examines calibration designs, which maximize the determinant of Fisher’s information matrix on the item parameters (D-optimal), for sets of polytomously scored items. These items were analyzed using a number of item response theory (IRT) models, which are members of the “divide-by-total” family, including the nominal categories model, the rating scale model, the unidimensional polytomous Rasch model and the partial credit model. We extend the known results for dichotomous items, both singly and in tests to polytomous items. The structure of Fisher’s information matrix is examined in order to gain insights into the structure of D-optimal calibration designs for IRT models. A theorem giving an upper bound for the number of support points for such models is proved. A lower bound is also given. Finally, we examine a set of items, which have been analyzed using a number of different models. The locally D-optimal calibration design for each analysis is calculated using an exact numerical and a sequential procedure. The results are discussed both in general and in relation to each other.

scholarly journals Beta Generalized Exponentiated Frechet Distribution with Applications

Open Physics ◽  
2019 ◽  
Vol 17 (1) ◽  
pp. 687-697
Author(s):  
Majdah M. Badr
Keyword(s):  
Maximum Likelihood ◽  
Hazard Rate ◽  
Information Matrix ◽  
Model Parameters ◽  
Fréchet Distribution ◽  
Method Of Maximum Likelihood ◽  
Frechet Distribution ◽  
Fisher's Information

Abstract In this article we introduce a new six - parameters model called the Beta Generalized Exponentiated-Frechet (BGEF) distribution which exhibits decreasing hazard rate. Many models such as Beta Frechet (BF), Beta ExponentiatedFrechet (BEF), Generalized Exponentiated-Frechet (GEF), ExponentiatedFrechet (EF), Frechet (F) are sub models. Some of its properties including rth moment, reliability and hazard rate are investigated. The method of maximum likelihood isproposed to estimate the model parameters. The observed Fisher’s information matrix is given. Moreover, we give the advantage of the (BGEF) distribution by an application using two real datasets

Sign in / Sign up

Export Citation Format

Share Document