COVID-19 Update: We are currently shipping orders daily. However, due to transit disruptions in some geographies, deliveries may be delayed. To provide all customers with timely access to content, we are offering 50% off Science and Technology Print & eBook bundle options. Terms & conditions.
Multivariate Analysis—III - 1st Edition - ISBN: 9780124266537, 9781483265131

Multivariate Analysis—III

1st Edition

Proceedings of the Third International Symposium on Multivariate Analysis Held at Wright State University, Dayton, Ohio, June 19-24, 1972

Editor: Paruchuri R. Krishnaiah
eBook ISBN: 9781483265131
Imprint: Academic Press
Published Date: 1st January 1973
Page Count: 428
Sales tax will be calculated at check-out Price includes VAT/GST
Price includes VAT/GST

Institutional Subscription

Secure Checkout

Personal information is secured with SSL technology.

Free Shipping

Free global shipping
No minimum order.


Multivariate Analysis — III contains the proceedings of the Third International Symposium on Multivariate Analysis held at Wright State University in Dayton, Ohio, on June 19-24, 1972. The papers explore the theory and applications of multivariate analysis and cover areas such as time series and stochastic processes; distribution theory and inference; characteristic functions and characterizations; and design and analysis of experiments. Classification, modeling, and reliability are also discussed. Comprised of 27 chapters, this volume begins with an introduction to two-dimensional random fields, giving results for a class of Gaussian processes with a multidimensional time parameter. The next chapter deals with concepts of consistency in spectral estimation for multivariate time series and considers the alternative of estimating the spectral distribution function or the spectral density function. Abstract martingales and ergodic theory are also examined, along with methods for assessing multivariate normality; inference and redundant parameters; characterization of the multivariate geometric distribution; and max-min designs in the analysis of variance. This monograph will be useful to statisticians and probabilists, as well as to scientists in other disciplines who are broadly interested in multivariate analysis.

Table of Contents

List of Contributors



Part I/Time Series and Stochastic Processes

Two-Dimensional Random Fields


Details of the Proof of Theorem 1


Concepts of Consistency in Spectral Estimation for Multivariate Time Series

1. Introductory Remarks

2. Should One Estimate the Spectral Distribution Function or Its Derivative?

3. Estimation of the Spectral Distribution Function F for a Multivariate Stationary Random Sequence



Non-Anticipative Canonical Representations of Equivalent Gaussian Processes

1. Introduction

2. The General Form of the Non-Anticipative Representation

3. A Derivation Using Martingale Theory

4. Concluding Remarks


Abstract Martingales and Ergodic Theory


1. The Problem

2. Martingale Formulation

3. An Operator Theoretic Approach

4. A Maximal Inequality

5. Final Remarks


On the Modelling and Estimation of Communication Channels

1. Introduction and Preliminary Discussion

2. Classes of Channels and Representations

3. Estimation of Parameters in a Linear Model

4. Channel Identification

5. Remarks


Innovation and Nonanticipative Processes

1. Innovation Processes and Regularity

2. Canonical Representations and Fully Submitted Processes


Part II/Distribution Theory and Inference

Methods for Assessing Multivariate Normality

1. Introduction

2. Univariate Techniques for Evaluating Marginal Normality

3. Multivariate Techniques for Evaluating Joint Normality

4. Tests Based on Unidimensional Views of Multivariate Data

5. Examples

6. Concluding Remarks


Asymptotic Expansions for the Distributions of Characteristic Roots When the Parameter Matrix Has Several Multiple Roots

1. Introduction

2. The Maximization Procedures

3. Asymptotic Expansion for the Distribution of the Latent Roots of the Estimated Covariance Matrix—Several Multiple Population Roots

4. Asymptotic Expansion for the Distribution of the Latent Roots of S1 S21—Several Multiple Population Roots

5. Asymptotic Expansion for Manova—Several Multiple Population Roots

6. Asymptotic Expansion for Canonical Correlation—Several Multiple Population Roots

7. Complex Analogues of Previous Results

8. Remarks


Aspects of the Multinomial Logit Model

1. The General Logit Model

2. Properties of the Likelihood

3. Comment

Appendix: The Beaton Sweep


Inference and Redundant Parameters

1. Introduction

2. The Probability Space Model

3. Measure Factorizations

4. If the Inner Parameter Become Known

5. A Redundant Parameter

6. The Multivariate Model

7. The Bayesian Right Invariant


The Variance Information Manifold and the Functions on It

1. The Variance Information Manifold and Its Boundary

2. The Bivariate Case

3. The Multinormal Distribution with Singular Information Matrix

4. Derivation via the Distribution of Linear Functions

5. Application to the Analysis of Experimental Designs

6. Representations as the Marginal Distribution of a Nonsingular Distribution

7. Decomposition of a Multinormal Distribution

8. Invariant Metric

9. Geodesic Distance between Two Matrices

10. Zonal Polynomials


Stopping Time in Sequential Samples from Multivariate Exponential Families

1. Introduction

2. The Main Theorem

3. Application to Examples 1.1 and 1.2


Part III/Characteristic Functions and Characterizations

An Isomorphism Method for the Study of I0n

1. Introduction

2. Definitions and Notations

3. Isomorphism Method

4. Applications in the General Case

5. Applications to a Finite Independent Set

6. Applications to An Enumerable Independent Set

7. Finite Products of Poisson Laws

8. α-Decompositions


A Characterization of the Multivariate Geometric Distribution

1. Introduction

2. A Regression Property

3. The Characterization Theorem

4. Derivation of the Differential Equations

5. Completion of the Proof of Sufficiency

6. Proof of the Necessity


On Infinitely Decomposable Probability Distributions, and Helical Varieties in Hilbert Space

1. Introduction

2. The Canonical Association of Helical Varieties with Infinitely Decomposable Distributions

3. A Hilbert Space Proof of the Lévy-Khinchine Theorem for R0

4. Operator-Measure Theoretic Treatment

5. On Probability and Hilbert Spaces

6. Bibliographical and Concluding Remarks


Limit Laws for Sequences of Normed Sums Satisfying Some Stability Conditions


Part IV/Design and Analysis of Experiments

The Analysis of Time Series Collected in an Experimental Design

1. Introduction

2. The Finite Fourier Transform

3. The Fixed Effects Model

4. The Random Effects Model

5. The Point Process Case



Max-Min Designs in the Analysis of Variance

1. Introduction and Max-Min Designs

2. A Matrix Inequality and Regular Designs


Analysis of Covariance Structures

1. Introduction

2. General Results

3. Applications


Optimum Designs for Fitting Biased Multiresponse Surfaces

1. Introduction

2. Formulation of Box and Draper

3. Minimizing V + B

4. An Illustrative Example

5. Other Comments


Asymptotic Properties of Some Sequential Nonparametric Estimators in Some Multivariate Linear Models

1. Introduction

2. The Problems

3. Preliminary Notions and Basic Assumptions

4. Asymptotic Properties of Robust Sequential Point Estimators of (α,β)

5. Bounded Length (Sequential) Confidence Bands for θ


Part V/Classification, Modelling, and Reliability

Availability Theory for Multicomponent Systems

Introduction and Summary

1. Preliminaries

2. Average System Up Time : Almost Sure Results

3. Asymptotic Distributions

4. Cost of Repair


Some Measures for Discriminating Between Normal Multivariate Distributions with Unequal Covariance Matrices

1. Summary and Introduction

2. The Measure S

3. The Measure T

4. The Kullback-Leibler Information Numbers


Correlation and Affinity in Gaussian Cases

1. Introduction

2. Correlation Coefficient and ρI ρII

3. Canonical Correlation and ρI or ρII


Identification of the Structure of Multivariable Stochastic Systems

1. Introduction

2. Two Results in Principal Components Analysis

3. Multivariable Linear Systems

4. Identification of the System Structure

5. Reduction of the Dimension of the Output Vector

6. Reduction of the Dimension of the Input Vector

7. Practical Problems

8. Tests of Significance of Eigenvalues

9. Testing the Equality of λk+1 (ω) = λk+2 (w) = ...λρ(ω) = λ(ω)

10. Asymptotic Theory for the Distribution of Eigenvalues


An Information Function Approach to Dimensionality Analysis and Curved Manifold Clustering

1. Introduction

2. Entropy in the Discrete and Continuous Cases

3. Uncertainty and Dimesionality

4. Some Auxiliary Techniques

5. Monte Carlo Studies


Nonlinear Iterative Partial Least Squares (NIPALS) Modelling: Some Current Developments

1. Introduction and Summary

2. What is NIPALS Modelling?

3. Low Information versus High Information Modelling

4. Causal Flow Models in Econometrics and the Behavioural Sciences

5. Case Studies in NIPALS Modelling


Titles of Contributed Papers


No. of pages:
© Academic Press 1973
1st January 1973
Academic Press
eBook ISBN:

About the Editor

Paruchuri R. Krishnaiah

Ratings and Reviews