The II International workshop on inductive modelling IWIM'2007

The II International workshop on inductive modelling (IWIM-2007)was held in Prague within the period from 23 to 26 of September, 2007. The first workshop took place in Kyiv in July 2005 as continuation of the International conference on inductive modelling (ICIM-2002) in Lviv in May 2002. Series of such conferences and workshops are international scientific forums, devoted to the theory, algorithms, applications, realization and new technologies of data analysis and extraction of knowledge based on methodology of the Group Method of Data Handling (GMDH) as a typically inductive method of modeling developed by A.G.Ivakhnenko.

Czech Technical University (CTU) in Prague and International Research and Training Center of Information Technologies and Systems (IRTC ITS) of NAN and MON of Ukraine were the organizers of the workshop.

Alexei Ivakhnenko, professor, academician of NAN of Ukraine, was the Honorary Chairman of the Workshop, the Workshop Chairman was Volodymyr Stepashko, professor, IRTC ITS, Ukraine, Co-chairman Godfrey Onwubolu, professor, South-Pacific University, Fiji, Scientific secretary Pavel Kordik, associate professor, CTU, Chairman of the Local Organizing Committee – Miroslav Snorek, professor, CTU. Members of the International Program Committee were scientists from Great Britain, Germany, Iran, Canada, China, USA, France, South Korea and Japan.

The workshop program cover such main topics: principles and theoretical bases of inductive modelling; optimization and evolution of inductive models and neuron networks; discovering of relationships and automation of knowledge extraction processes; analysis and prediction of time series; parallel and distributed computing; visual data analysis; applications to real world problems, including ecology, sociology, medicine, bioinformatics etc.

The Program of IWIM-2007 included 55 papers covered the wide range of questions of the theory, technologies and applications of inductive approach to the problems of modelling and prediction with the purpose of decision making in economy, ecology, medicine and other areas. Papers were distributed by countries as follows: 24 papers from Ukraine, 15 from Czech Republic, 5 from Republic Fiji, by 2 from Spain and Japan and by one from Great Britain, India, Iran, China, Russia, France and South Korea.

At the first day of the Workshop work, main tutorials and plenary lectures were presented, at the second and third days section and poster sessions, at the final day plenary lectures, general discussion and closing of the Workshop took place.

The follows tutorials were presented:
1. Volodymyr Stepashko (Kyiv, Ukraine). Inductive modelling on the basis of GMDH: theoretical, structural and applied aspects (in two parts).
2. Frank Lemke (Berlin, Germany). Multi-leveled Self-organization.
3. Pavel Kordik (Prague, Czech). GMDH and FAKE GAME: Evolving ensembles of inductive models.

Titles of the presented plenary papers:
1. Tadashi Kondo (Tocushima, Japan). Revised GMDH-Type Neural Network algorithms with Self-Selecting Optimum Architecture and Their Applications to Nonlinear System Identification and 3-Dimensional Medical Image. Recognition.
2. Vladimir Lytvynenko (Kherson, Ukraine). Theoretical and applied aspects of the hybrid immune systems development.
3. Godfrey Onwubolu (Suva, Fiji). Designing Hybrid Computational Intelligence Methodologies and Group Method of Data Handling for Inductive Modeling of Nonlinear Systems.
4. Tetyana Aksenova (Grenoble, France). GMDH-PNN and its applications in medicine.

Papers of other workshop participants were presented in the sections:
1) principles and theoretical bases of inductive modelling;
2) optimum complexity of inductive models;
3) optimization and evolution of inductive models and neuron networks;
4) data mining;
5) analysis and prediction of time series;
6) application of inductive modelling to real world problems.

Within the framework of Section 1 one should point out such papers:
P.Buryan (Czech), «Enhanced MIA GMDH algorithm»;
S.Zhyliaev, D.Ryabokon (Ukraine), «Optimization of the Bayesian classifier's structure using GMDH algorithm to forecast Internet-clients' preferences»;
L.Sarycheva (Ukraine), «The spatial-temporary approach in problems of clusterization»;
N.Kondrashova (Ukraine), «Influence of sample division on the quality of modeling and forecasting of real processes»;
S.Yefimenko, V.Stepashko (Ukraine), «Computer tests as an instrument for effectiveness investigation of modeling algorithms» is.

The most interesting papers of Section 2:
A.Ivakhnenko, E.Savchenko, L.Somina (Ukraine), «Inductive Method of Optimal Model Selection by External Error Criterion with Additional Determination Using Bias»;
M.Dyvak, V.Manzhula, A.Pukas (Ukraine), «Structural identification of interval models of the static systems»;
O.Sarychev (Ukraine), «Modelling in the Class of Regression Equations Systems»;
O.Koshulko, A.Koshulko (Ukraine), «Adaptive parallel implementation of the Combinatorial GMDH algorithm».

Within the Section 3 one may note such papers:
G.Ivahnenko (England), «Marketing Problems Solution by Different GMDH Algorithms Using Excel Software»;
G.Onwubolu (Fiji), «Data Mining using Inductive Modeling Approach»;
J.Drhal, M.Snorek, P.Kordik (Czech), «Dataset visualization based on a simulation of intermolecular forces».

Within the Section 4 it is needed to pay attention to papers:
O.Samoylenko, V.Stepashko (Ukraine), «Combinatorial GMDH algorithm with successive selection of arguments»;
N.Nariman-zade, A.Janiali (Iran), the «Pareto Genetic Design of GMDH-type Neural Networks for Nonlinear Systems»;
M.Jirina, M.Jirina jr. (Czech), «Genetic Selection and Cloning in GMDH MIA Method ».

Within the framework of Section 5 the most interesting there were papers:
V.Lytvynenko, P.Bidyuk, J.Bardachov at al. (Ukraine), «The Combined Immune Algorithm Based on Clonal Selection»;
J.Koutnik (Czech), «Inductive Modelling of Temporal Sequences by Means of Self-organization»;
J.Bouska, P.Kordik (Czech), «Time Series Prediction by means of GMDH Analog Complexing and GAME».

The work of Section 6 was the most dense, where a wide spectrum of applications was presented:
M.Cepek, V.Chudadek, M.Petrik at al. (Czech), «Comparison of Inductive Modeling Method to Other Classification Methods for Holter ECG»;
V.Vissikirsky, V.Stepashko (Ukraine), I.Kalavrouziotis (Greece), «Application of GMDH to the Environmental Modeling with Short Samples»;
N.Kondrashova, A.Pavlov, Ya.Pavlov (Ukraine), «Optimization of Forecasting Models for Testing Blood Samples by Estimation of Tiol-Disulfid Diagrams»;
A.Pilny, P.Kordik (Czech), «Reconstruction of Eye Movements Signal using Inductive Model Detecting Saccades»;
D.Nowicki, V.Shaposhnik, A.Bouamrani at al. (France), «GMDH-based Approach for Analysis of Mass Spectra in Clinical Proteomics»;
V.Gerla, M.Boursa, L.Lhotska at al. (Czech), «Inductive Modeling in Newborn Sleep Stage Recognition» etc.

All the papers of the Workshop have published in Proceedings of the Prague Workshop [1] available in the electronic form at the proper web-site [2], and also at the traditional GMDH site [3].

1. Proceedings International Workshop on Inductive Modelling (IWIM 2007). – Prague: Czech Technical University, 2007. – 329 p. – ISBN 978-80-01-03881-9.
2. http://iwim2007.felk.cvut.cz
3. GMDH (http://www.gmdh.net).