PROCEEDINGS OF THE FIRST US/JAPAN CONFERENCE ON THE FRONTIERS OF STATISTICAL MODELING

 

VOLUME 3: ENGINEERING AND SCIENTIFIC

APPLICATIONS

 

EDITED BY

H. BOZDOGAN

 

KLUWER ACADEMIC PUBLISHERS 1994

DORDRECHT, THE NETHERLANDS

 

Ordering Address:

E-mail: Kluwer@world.std.com

 

® Designed by H. Bozdogan & Created by Hugh Bailey, 1991.

 
PREFACE TO VOLUME 3
ISBN 0-7923-2599-0 (Volume 3)
This is the third volume of the proceedings of The US/Japan Conference on the Frontiers of Statistical Modeling: An Informational Approach. Papers included in this volume deal with engineering and scientific applications.
As a result of Akaike's (1973) AIC, many important statistical modeling techniques have been developed in engineering and scientific applications. Since 1987 in Japan, many supercritical thermal power plants are governed by dynamic statistical controllers. On line multivariate autoregressive time series models are used to maximize the capacity of the production of electricity and to minimize the cost of production. Other "Informational Modeling" techniques have been used to control ship's motions and marine engines for energy conservation, in studying prediction and occurrences of earthquakes, and in data analysis taken aboard a spinning spacecraft such as the Pioneer Venus for signal extraction.
The first paper is another keynote lectures given by Professor Akaike and discusses the implications of the informational viewpoint for statistical science. The rest of the papers included in this volume deal with the applications of model selection and evaluation in optimal control of thermal power plants, seismological applications, multichannel studies, nonlinear filter designs, inference of evolutionary trees from DNA and protein sequence data, group method of data analysis, neural binary classification, studying directional interactions, and stable estimation techniques.
It is with great pleasure and gratitude that the editor and co-editors acknowledge the authors and the referees who contributed to this volume, and the chairs of this scientific core of the conference.
 
Hamparsum Bozdogan
The University of Tennessee
Knoxville, July 1993 
CONTENTS OF VOLUME 3

 List of Contributors to Volume 3  vii
 Editor's General Preface  ix
 Preface to Volume 3  xiii

Summary of Contributed Papers to Volume 3



  
1. Implications of Informational Point of View on the Development
  of Statistical Science (Keynote lecture) 
  
H. Akaike 


 27



  
2. From Comparison Density to Two Sample Analysis 
  
E. Parzen 


 39

  
3. Statistical Identification and Optimal Control of Thermal
  Power Plants 
  
H. Nakamura 


 57



  
4. Applications of Auto Regressive Model to Control Ship's
  Motions and Marine Engine 
  
K. Ohtsu and G. Kitagawa 


 81

  
5. Statistical Models for Earthquake Occurrence: Clusters,Cycles
  and Characteristic Earthquakes 
  
D. Vere-Jones 


 105


  
6. Seismological Applications of Statistical Methods for
  Point-Process Modelling 
  
Y. Ogata 


 137

  
7. One Channel At-A-Time Multichannel Autoregressive Modeling
  of Stationary and Nonstationary Time Series 
  
W. Gersch and D. Stone 


 165


  
8. Separation of Spin Synchronized Signals Using a Bayesian
  Approach 
  
T. Higuchi 


 193

  
9. The Local Linearization Filter With Application to Nonlinear
  System Identifications 
  
T. Ozaki 


 217


  
10. Inference of Evolutionary Trees From DNA and Protein
  Sequence Data 
  
M. Hasegawa   


 241

  
11. New Structure Criteria in Group Method of Data Handling
  
  
T. Lange 


 249


  
12. The Use of the Kullback-Leibler Distance for Learning
  in Neural Binary Classifiers 
  
D.Pados, P.Papantoni-Kazakos, D.Kazakos, and A.Koyiantis 


 267

  
13. Likelihood Estimation of Directional Interaction
  
  
M. Tanemura 


 293


  
14. New Approach to Optimization of Stable Estimation 
  
A. M. Shurygin 


 315