An Information Theoretic Approach to Econometrics by George G. Judge

By George G. Judge

This booklet is meant to supply the reader with an organization conceptual and empirical knowing of uncomplicated information-theoretic econometric versions and strategies. simply because such a lot facts are observational, practitioners paintings with oblique noisy observations and ill-posed econometric versions within the type of stochastic inverse difficulties. for this reason, conventional econometric equipment in lots of situations are usually not acceptable for answering a few of the quantitative questions that analysts desire to ask. After preliminary chapters take care of parametric and semiparametric linear chance types, the focal point turns to fixing nonparametric stochastic inverse difficulties. In succeeding chapters, a relatives of energy divergence measure-likelihood features are brought for a number conventional and nontraditional econometric-model difficulties. eventually, inside of both an empirical greatest probability or loss context, Ron C. Mittelhammer and George G. pass judgement on recommend a foundation for selecting a member of the divergence relations

Show description

Read Online or Download An Information Theoretic Approach to Econometrics PDF

Similar econometrics books

Happiness Quantified: A Satisfaction Calculus Approach

How will we degree happiness? this significant and long-awaited publication provides a brand new and unified method of the research of subjective delight and source of revenue overview. Drawing on empirical analyses of German, British, Dutch, and Russian info, it develops new method to set up a version of health and wellbeing including pride with existence as a complete and with a variety of domain names of existence.

Market Response Models: Econometric and Time Series Analysis

From 1976 to the start of the millennium—covering the quarter-century lifestyles span of this publication and its predecessor—something awesome has occurred to marketplace reaction study: it has develop into perform. lecturers who educate in expert fields, like we do, dream of such issues. think the pride of figuring out that your paintings has been integrated into the decision-making regimen of brand name managers, that class administration will depend on recommendations you built, that advertising and marketing administration believes in whatever you struggled to set up of their minds.

Allocation Models and their Use in Economic Planning

3 assorted traces of process have contributed to the speculation of optimum making plans. One method considers the matter from the view-point of a countrywide govt and its adviser, the econometrician making plans speci­ alist. the govt can, if this can be regarded as fascinating, stimulate funding in yes instructions and discourage different financial actions.

Mastering 'Metrics: The Path from Cause to Effect

Utilized econometrics, identified to aficionados as 'metrics, is the unique facts technological know-how. 'Metrics encompasses the statistical equipment economists use to untangle reason and impact in human affairs. via obtainable dialogue and with a dose of kung fu-themed humor, learning 'Metrics offers the basic instruments of econometric examine and demonstrates why econometrics is interesting and important.

Extra info for An Information Theoretic Approach to Econometrics

Sample text

In this fully parametric model formulation, a large amount of information about the sampling process is specified but the parameters β and σ 2 nonetheless remain unknown and unobserved and are, indeed, unobservable. 10) and Because the ML estimator for β is a linear combination of the multivariate normally distributed vector Y, it follows that the ML estimator has a normal distribution N(β, σ 2 (x x)−1 ). It can also be shown that the ML estimator of n σ 2 can be bias-adjusted to define the unbiased estimator S2 = ( n−k )S2ML ∼ n−k 2σ2 Gamma ( 2 , n−k ).

The unbiasedness property follows from the condition E [ε |X] = 0, and E [βˆ ] = E x [E [βˆ |X]] = E x [β + (X X)−1 X E [ε|X]] = β. 10) given that E [(X X)−1 ] exists. 11) and, therefore, the estimator S 2 is a conditionally and unconditionally unbiased estimator of σ2 . 3 LS Estimator Properties: Asymptotics The asymptotic properties of the LS estimator when X is stochastic parallel those established in the case where x is fixed. 4 General Linear Model with Stochastic X 27 p p estimator, if n−1 X ε → 0 and n−1 X X → , then βˆ exists with probability p converging to 1 and βˆ → β when n → ∞.

This result is precisely the same as the case of nonstochastic x. The typical MLE asymptotics of consistency, asymptotic normality, and asymptotic efficiency also apply to these estimators. Furthermore, βˆ is conditionally n 2 normally distributed in finite samples, and S 2 = ( n−k )S ML is conditionally Chi-square(n-k, 0) distributed. 3. The asymptotic distributions of the test statistics are again Chi-square (W, LM, and pseudo-likelihood ratio [PLR]statistics) and normal (Z-statistic). 3. These assumptions are sufficient conditions.

Download PDF sample

Rated 4.85 of 5 – based on 26 votes