There are more than 500 on-line documents connected with the GMDH in Internet (in January 2001). They can be divided into books, publications, conference thesisis, industrial projects, software, doctoral thesisis, university programs, mailing lists archives. The most part are investigations from Japan, but unfortunately they have not english translation.



  • Books
  • Articles. References
  • Conference thesisis
  • Projects
  • Software
  • Programs and Thesisis

    There are some of interesting GMDH links
    ( - stands for dead link and - seems to be temporary unavailable):


    1. Intelligent Systems Laboratory, UK: D.T. Pham, X. Liu Neural Networks for Identification, Prediction and Control - Springer-Verlag NY, 1995, 238p. ISBN 3-540-19959-4. Monograph.
      (alternate1) (alt2)
    2. Eric Rogers (ed.) Parallel processing in a control systems environment. Series in Systems and Control Engineering., Prentice-Hall International London, 1993, 364p., ISBN 0-13-651530-4
    3. Hecht-Nielsen,R., Neurocomputing, Addison Wesley, 1990.  
    4. Reevel/Zenith (ed.) Parallel Processing and Artificial Intelligence, Wiley&Sons, 1989. 
    5. Computational Finance Bibliography Updated 11.95. ...GMDH is a scheme for fitting a high-degree sparse polynomial to data. As such, it can do some jobs that neural nets also do. This scheme, also called "abductive modeling" by some, is a type of regression without models. This book is a collection of papers on...

    Articles and Abstracts

    • Kuo-lin Hsu1, Hoshin Vijai Gupta, and Soroosh Sorooshian. Artificial neural network modeling of the rainfall-runoff process. Department of Hydrology and Water Resources, The University of Arizona, Published in Water Resources Research, 29 (4), pp. 1185-1194, April 1993
    • Tadashi Kondo Multiinput-Multioutput Type GMDH Algorithm Using Regression Component Analysis. School of Medical Sciences, University of Tokushima, Japan. Transactions of Systems, Control and Information, Vol. 6, No. 11 (1993). The GMDH algorithm in this paper uses total characteristic variables, which can explain variation of all output variables, and optimal partial polynomials are constructed by combining these total characteristic variables.
    • Tadashi Kondo Identification of Large-Spatial Air Pollution Patterns Using a Neural Network. School of Medical Sciences, University of Tokushima, Japan. Transactions of systems, control and information vol.7, no.2 (1994).
    • Tadashi Kondo Multiinput-Multioutput Type GMDH Algorithm Using Regression Component Analysis. School of Medical Sciences, University of Tokushima, Japan. Transactions of Systems, Control and Information, Vol. 6, No. 11 (1993). The GMDH algorithm in this paper uses total characteristic variables, which can explain variation of all output variables, and optimal partial polynomials are constructed by combining these total characteristic variables.
    • Lee,Shin-Ping, Chang,Fi-John Study on Reservoir Operation By Using Fuzzy Stochastic Dynamic Programming. Taiwan College. A back-propagation of ANN and GMDH have been applied for investigating the hourly rainfall intensity with different typhoon event in Hua-Lien area. At last, the method that combines of rainfall forecasting and fuzzy control theory to forecast the degree of debris flow is simulated. The results show that the method can make an excellent judgement of the trend of debris flow.
    • Article Neural Networks and Currency Exchange Rate Prediction by John G. Carney and Dr. Padraig Cunningham
    • Hayashi Lab. in Hannan University, Japan.
      (On Three Formulations of Fuzzy Linear Regression Analysis) Vol.22, No.10, pp.1051-1057 (1986))
      (The Formulation of the Fuzzy GMDH) Vol.30, No.9, pp.581-587 (1986)) (Analysis and Prediction of Water Temperatures in a Reservior by the Fuzzy GMDH) (Vol.23, No.12, pp.1304-1311 (1987)
      (The Fuzzy GMDH Algorithm by Possibility Models and Its Application Isao Hayashi, Hideo Tanaka Fuzzy Sets and Systems, Vol.36, No.2, pp.245-258 (1990)
    • Intelligent Systems Laboratory, UK. D.T.Pham and X.Liu - 'Modelling and prediction using GMDH networks of Adalines with nonlinear preprocessors', Int J Systems Sci, Vol.25, No.11 (1994)
    • Professor David G. Green, AU. - Publications 33.Green, D.G., Reichelt, R.E., and Bradbury, R.H. (1988). Statistical behaviour of the GMDH algorithm. Biometrics 44, 49-69.
    • Publications of Michio J. Kishi Kishi, M.J. and N.Suginohara: Effects of longshore variation of coastline geometry and bottom topography on coastal upwelling...
    • DMS publications, AU. Stone, R. Title : Cautions about GMDH - type modeling of data from manufacturing processes. Publication: Journal of Manufacturing Systems
    • Oceanographic literature review Volume 42, Issue 6 Tables of Contents service. June 1995
    • Abstracts of Research Reports available from Elegant Mathematics, Inc. The Research Reports described below contain most of the mathematical algorithms.  
    • Stephen Grossberg and John W. L. Merrill Formal and Biological Neural Networks for Adaptively Timed...
    • Operations Research Program at the University of Delaware Ph.D. investigation of R. Nanda: M.S., 1991. "Application of Fuzzy Set Theory and Neutral Networks to Regression and GMDH Analysis", Dr. Shinya Kikuchi/Civil Engineering, advisor. University of Delaware - Ph.D. study Professor P.Krishnan
    • S B Hasnain and D A Linkens, The use of transputer parallelism for the group method of data handling (GMDH) self-organising idenification algorithm, IEEE Workshop: Parallel processing in control -- the transputer and other architectures, Bangor, 1987, 77. (alternate)

    Conferences thesisis

    • IV Simposio Nacional de Prediccion Memorial "Alfonso Ascaso" Madrid 15-19 Abril 1996. Aactualizacion de la metodologia de prediccion de temperaturas extremas en baleares. José Antonio Guijarro Pastor. De las contribuciones aparecidas en revistas extranjeras citaremos únicamente dos muy recientes: 2) Abdel-Aal y Elhadidy (Abdel-Aal, R.E., Elhadidy, M.A., 1995. Modeling and Forecasting the Daily Maximum Temperature Using Abductive Machine Learning. Weather and Forecasting, vol.10, pp. 310-325, 1995), que aplican un modelo neuronal basado en las técnicas GMDH, aunque usando únicamente datos de superficie. Todavía se intentó generar un quinto modelo, basado en una red neuronal del tipo GMDH, pero no se pudo construir ninguna que mejorara los resultados de los modelos de regresión lineal.
    • 4th IEEE International Conference on Fuzzy Systems & 2nd International Fuzzy Engineering Symposium, held in Yokohama Japan, 20-24 March 1995. FUZZ-IEEE/IFES'95
      Selection of the Optimum Number of Hidden Layers in Neuro-Fuzzy GMDH H.Ichihashi, N. Harada and K. Nagasaka - University of Osaka Prefecture,
      Motion planning of a redundant manipulator - criteria of skilled operators by fuzzy-ID3 and GMDH and optimization by GA. Shibata T., Abe T., Tanie K., and Nose M. Mechanical Engineering Laboratory and Komatsu Ltd. This paper proposes a modeling method of criteria of skilled operators for motion planning of a redundant manipulator in industrial applications. The method employs Fuzzy-ID3 to extract important factors from the criteria and GMDH to model them to evaluate motion plans. Experimental results show the effectiveness of the proposed method. (alternate1)
    • 9th Fuzzy System Symposium (Sapporo Japan 19-21 May 1993) Neuro-Fuzzy GMDH and its Application to Modeling of Grinding Characteristics Kazunori Nagasaka (College of Engineering, University of Osaka Prefecture), et al.
      Abstract:Fuzzy modeling is presented for reducing the number of production rules in expert system. GMDH is used for identification of the model having such many input variables with small number of data by using hierarchical structure. This paper treats the proposition of neuro-fuzzy GMDH algorithm adopting fuzzy models as partial descriptions in conventional GMDH.
      GMDH by Fuzzy If-Then Rules Hideo Tanaka (University of Osaka Prefecture), et al.
      Abstract: A method of automatic learning of fuzzy if-then rules with certainty factors from the given input-output data is developed.
    • Voordrachten en Colloquia W&I NL, Bulletin 01.1995 Onderzoekscolloquium Vakgroep Informatiesystemen Drs. J. van der Vlugt, Toepassing van de Group Method of Data Handling voor het vinden van Leaders-Followers effecten op de Amsterdamse beurs. Samenvatting: De afdeling Informatie Technologie Research van de ING Bank onderzoekt o.a. de mogelijkheden en de bruikbaarheid van diverse methoden en technieken voor het voorspellen van beurskoersen. Gezien de grote aantallen berekeningen die moeten worden uitgevoerd (80.000.000 regressies over gemiddeld zo'n 2000 datapunten) zijn deze op een 32-processor nCUBE uitgevoerd.
    • CSNA'95 Program Abstracts Classification Society of North America 1995 annual meeting. Colorado, June 22-25, 1995.
      C.Hild & H. Bozdogan, University of Tennessee, Tree-based classification of predictors in regression models and structure determination via self-organization: the GMDH...
      ...method called the Ivakhnenko (1968) Polynomials of Several Predictors or the Group Method of Data Handling (GMDH). The GMDH is an elegant approach to statistical data modeling to determine the relationships among the m given predictors and to describe the structure of high-order regression type models...
    • DAAAM International Symposium 1995, Poland
      Jaeckel, O. and Härtig, G. GMDH-Algorithmen als Mittel zur Strukturoptimierung von Polynom-Modellen (alternate)
    • DAAAM Symposium 1996 TU Wien Product & Manufacturing: Flexibility, Integration, Intelligence
      Jaeckel, O. and Härtig, G. Funktionsweise und Weiterentwicklung von GMDH-Algorithmen
    • ANNIE'96 Artificial neural networks in engineering 1996, St.Louis, Missouri Fusion of Fuzzy and Chaos, and its Application in Industrial Field: Time Series Prediction Tadashi Iokibe, Meidensha Corporation, Tokyo, JAPAN
    • No Title. There are 4 abstracts for the conference PIERS 1996. These abstacts are given.


    • NOx Emissions Modeling and Prediction for a Gas Turbine. Sponsor: General Electric, Drive Systems Division, Roanoke, VA. Advisors: G. Vachtsevanos, and Doug McCallum. Dates: March 1994 - March1995.
    • IMCSL Projects The objective of this project is to design a high-level intelligent controller that will master the low-level flight controllers for the Autonomous Scout Rotorcraft Testbed helicopter with the following autonomous capabilities: Route planning and tracking ...
    • Institutsveranstaltungen, 1993. (ISWW) - Karlsruher Flockungstage. Diese vom Institut für Siedlungswasserwirtschaft jährlich veranstaltete Tagung stellt ein... Projektberichte 1993 (ISWW) Ecological Management of Municipal Wastewater in Shenyang, China. (alternate)
    • Projektforschung - Humboldt-Universität zu Berlin Institut für Informatik. Jahresbericht 1995. Lehr-und Forschungseinheit. Automaten-und...  


    Programs and Thesisis