Group method of data handling

Group method of data handling (GMDH) is a family of inductive algorithms for computer-based
mathematical modeling of multi-parametric datasets that features fully automatic structural and parametric
optimization of models.

GMDH is used in such fields as data mining, knowledge discovery, prediction, complex systems modeling,
optimization and pattern recognition.[1] GMDH algorithms are characterized by inductive procedure that
performs sorting-out of gradually complicated polynomial models and selecting the best solution by means
of the external criterion.

A GMDH model with multiple inputs and one output is a subset of components of the base function (1):

where fi are elementary functions dependent on different sets of inputs, ai are coefficients and m is the
number of the base function components.

In order to find the best solution, GMDH algorithms consider various component subsets of the base
function (1) called partial models. Coefficients of these models are estimated by the least squares method.
GMDH algorithms gradually increase the number of partial model components and find a model structure
with optimal complexity indicated by the minimum value of an external criterion. This process is called
self-organization of models.

As the first base function used in GMDH, was the gradually complicated Kolmogorov–Gabor polynomial
(2):

Usually more simple partial models with up to second degree functions are used.[1]

The inductive algorithms are also known as polynomial neural networks. Jürgen Schmidhuber cites
GMDH as one of the first deep learning methods, remarking that it was used to train eight-layer neural nets
as early as 1971.[2][3]

History
The method was originated in 1968 by Prof. Alexey G. Ivakhnenko in the Institute of Cybernetics in Kyiv.
This inductive approach from the very beginning was a computer-based method so, a set of computer
programs and algorithms were the primary practical results achieved at the base of the new theoretical
principles. Thanks to the author's policy of open code sharing the method was quickly settled in the large
number of scientific laboratories worldwide. As most routine work is transferred to a computer, the impact
of human influence on the objective result is minimised. In fact, this approach can be considered as one of
the implementations of the Artificial Intelligence thesis, which states that a computer can act as powerful
advisor to humans.
The development of GMDH consists of a synthesis of ideas from different
areas of science: the cybernetic concept of "black box" and the principle of
successive genetic selection of pairwise features, Godel's incompleteness
theorems and the Gabor's principle of "freedom of decisions choice",[4] the
Adhémar's incorrectness and the Beer's principle of external additions.[5]

GMDH is the original method for solving problems for structural-parametric

identification of models for experimental data under uncertainty.[6] Such a
problem occurs in the construction of a mathematical model that approximates
the unknown pattern of investigated object or process.[7] It uses information
about it that is implicitly contained in data. GMDH differs from other methods
of modelling by the active application of the following principles: automatic GMDH author – Soviet
models generation, inconclusive decisions, and consistent selection by external scientist Prof. Alexey G.
criteria for finding models of optimal complexity. It had an original Ivakhnenko.
multilayered procedure for automatic models structure generation, which
imitates the process of biological selection with consideration of pairwise
successive features. Such procedure is currently used in Deep learning networks.[8] To compare and choose
optimal models, two or more subsets of a data sample are used. This makes it possible to avoid preliminary
assumptions, because sample division implicitly acknowledges different types of uncertainty during the
automatic construction of the optimal model.

During development was established an organic analogy between the problem of constructing models for
noisy data and signal passing through the channel with noise.[9] This made possible to lay the foundations
of the theory of noise-immune modelling.[6] The main result of this theory is that the complexity of optimal
predictive model depends on the level of uncertainty in the data: the higher this level (e.g. due to noise) -
the simpler must be the optimal model (with less estimated parameters). This initiated the development of
the GMDH theory as an inductive method of automatic adaptation of optimal model complexity to the level
of noise variation in fuzzy data. Therefore, GMDH is often considered to be the original information
technology for knowledge extraction from experimental data.

Period 1968–1971 is characterized by application of only regularity criterion for solving of the problems of
identification, pattern recognition and short-term forecasting. As reference functions polynomials, logical
nets, fuzzy Zadeh sets and Bayes probability formulas were used. Authors were stimulated by very high
accuracy of forecasting with the new approach. Noise immunity was not investigated.

Period 1972–1975. The problem of modeling of noised data and incomplete information basis was solved.
Multicriteria selection and utilization of additional priory information for noiseimmunity increasing were
proposed. Best experiments showed that with extended definition of the optimal model by additional
criterion noise level can be ten times more than signal. Then it was improved using Shannon's Theorem of
General Communication theory.

Period 1976–1979. The convergence of multilayered GMDH algorithms was investigated. It was shown
that some multilayered algorithms have "multilayerness error" – analogous to static error of control systems.
In 1977 a solution of objective systems analysis problems by multilayered GMDH algorithms was
proposed. It turned out that sorting-out by criteria ensemble finds the only optimal system of equations and
therefore to show complex object elements, their main input and output variables.

Period 1980–1988. Many important theoretical results were received. It became clear that full physical
models cannot be used for long-term forecasting. It was proved, that non-physical models of GMDH are
more accurate for approximation and forecast than physical models of regression analysis. Two-level
algorithms which use two different time scales for modeling were developed.
Since 1989 the new algorithms (AC, OCC, PF) for non-parametric modeling of fuzzy objects and SLP for
expert systems were developed and investigated.[10] Present stage of GMDH development can be
described as blossom out of deep learning neuronets and parallel inductive algorithms for multiprocessor
computers.

External criteria
External criterion is one of the key features of GMDH. Criterion describes requirements to the model, for
example minimization of Least squares. It is always calculated with a separate part of data sample that have
not been used for estimation of coefficients. This makes it possible to select a model of optimal complexity
according to the level of uncertainty in input data. There are several popular criteria:

Criterion of Regularity (CR) – Least squares of a model at the sample B.

Criterion of Minimum bias or Consistency – a squared error of difference between the
estimated outputs (or coefficients vectors) of two models developed on the basis of two
distinct samples A and B, divided by squared output estimated on sample B. Comparison of
models using it, enables to get consistent models and recover a hidden physical law from
the noisy data.[1]
Cross-validation criteria.

A simple description of model development using GMDH

For modeling using GMDH, only the selection criterion and maximum model complexity are pre-selected.
Then, the design process begins from the first layer and goes on. The number of layers and neurons in
hidden layers, model structure are determined automatically. All possible combinations of allowable inputs
(all possible neurons) can be considered. Then polynomial coefficients are determined using one of the
available minimizing methods such as singular value decomposition (with training data). Then, neurons that
have better external criterion value (for testing data) are kept, and others are removed. If the external
criterion for layer's best neuron reach minimum or surpasses the stopping criterion, network design is
completed and the polynomial expression of the best neuron of the last layer is introduced as the
mathematical prediction function; if not, the next layer will be generated, and this process goes on.[11]

GMDH-type neural networks

There are many different ways to choose an order for partial models consideration. The very first
consideration order used in GMDH and originally called multilayered inductive procedure is the most
popular one. It is a sorting-out of gradually complicated models generated from base function. The best
model is indicated by the minimum of the external criterion characteristic. Multilayered procedure is
equivalent to the Artificial Neural Network with polynomial activation function of neurons. Therefore, the
algorithm with such an approach usually referred as GMDH-type Neural Network or Polynomial Neural
Network. Li showed that GMDH-type neural network performed better than the classical forecasting
algorithms such as Single Exponential Smooth, Double Exponential Smooth, ARIMA and back-
propagation neural network.[12]

Combinatorial GMDH
Another important approach to partial models consideration that becomes more and more popular is a
combinatorial search that is either limited or full. This approach has some advantages against Polynomial
Neural Networks, but requires considerable computational power and thus is not effective for objects with a
large number of inputs. An important achievement of
Combinatorial GMDH is that it fully outperforms linear regression
approach if noise level in the input data is greater than zero. It
guarantees that the most optimal model will be founded during
exhaustive sorting.

Basic Combinatorial algorithm makes the following steps:

Divides data sample at least into two samples A and B.

Generates subsamples from A according to partial
models with steadily increasing complexity.
Estimates coefficients of partial models at each layer of Fig.1. A typical distribution of
models complexity. minimal values of criterion of
Calculates value of external criterion for models on regularity for Combinatorial GMDH
sample B. models with different complexity.

Chooses the best model (set of models) indicated by

minimal value of the criterion.
For the selected model of optimal complexity recalculate coefficients on a whole data
sample.

In contrast to GMDH-type neural networks Combinatorial algorithm usually does not stop at the certain
level of complexity because a point of increase of criterion value can be simply a local minimum, see Fig.1.

Algorithms
Combinatorial (COMBI)
Multilayered Iterative (MIA)
GN
Objective System Analysis (OSA)
Harmonical
Two-level (ARIMAD)
Multiplicative–Additive (MAA)
Objective Computer Clusterization (OCC);
Pointing Finger (PF) clusterization algorithm;
Analogues Complexing (AC)
Harmonical Rediscretization
Algorithm on the base of Multilayered Theory of Statistical Decisions (MTSD)
Group of Adaptive Models Evolution (GAME)

List of software
FAKE GAME Project (https://web.archive.org/web/20080213145150/http://neuron.felk.cvut.c
z/game/project.html) — Open source. Cross-platform.
GEvom (https://web.archive.org/web/20080418084252/http://research.guilan.ac.ir/gevom/)
— Free upon request for academic use. Windows-only.
GMDH Shell (https://gmdhsoftware.com/predictive-analytics-software) — GMDH-based,
predictive analytics and time series forecasting software. Free Academic Licensing and Free
Trial version available. Windows-only.
 KnowledgeMiner (http://www.knowledgeminer.eu/about.html) — Commercial product. Mac
OS X-only. Free Demo version available.
PNN Discovery client (http://pnn.pnnsoft.com/index.html) — Commercial product.
Sciengy RPF! (https://sourceforge.net/projects/sciengyrpf/) — Freeware, Open source.
wGMDH (http://wgmdh.irb.hr/en/project/) — Weka plugin, Open source.
R Package (https://cran.r-project.org/web/packages/GMDH/) – Open source.
R Package for regression tasks (https://CRAN.R-project.org/package=GMDHreg) – Open
source.
Python library of MIA algorithm (https://github.com/kvoyager/GmdhPy/) - Open source.

References
1. Madala, H.R.; Ivakhnenko, O.G. (1994). Inductive Learning Algorithms for Complex Systems
Modeling (https://web.archive.org/web/20171231104312/http://articles.gmdh.net/theory/GMD
Hbook.zip). Boca Raton: CRC Press. ISBN 978-0849344381. Archived from the original (htt
p://articles.gmdh.net/theory/GMDHbook.zip) on 2017-12-31. Retrieved 2019-11-17.
2. Schmidhuber, Jürgen (2015). "Deep learning in neural networks: An overview". Neural
Networks. 61: 85–117. arXiv:1404.7828 (https://arxiv.org/abs/1404.7828).
doi:10.1016/j.neunet.2014.09.003 (https://doi.org/10.1016%2Fj.neunet.2014.09.003).
PMID 25462637 (https://pubmed.ncbi.nlm.nih.gov/25462637). S2CID 11715509 (https://api.s
emanticscholar.org/CorpusID:11715509).
3. Ivakhnenko, Alexey (1971). "Polynomial theory of complex systems" (http://gmdh.net/articles/
history/polynomial.pdf) (PDF). IEEE Transactions on Systems, Man and Cybernetics. SMC-1
(4): 364–378. doi:10.1109/TSMC.1971.4308320 (https://doi.org/10.1109%2FTSMC.1971.43
08320).
4. Gabor, D. (1971). Perspectives of Planing. Organization of Economic Cooperation and
Development. London: Imp.Coll.
5. Beer, S. (1959). Cybernetics and Management. London: English Univ. Press.
6. Ivakhnenko, O.G.; Stepashko, V.S. (1985). Pomekhoustojchivost' Modelirovanija (Noise
Immunity of Modeling) (https://web.archive.org/web/20171231104218/http://articles.gmdh.ne
t/theory/bookNoiseIm.pdf) (PDF). Kyiv: Naukova Dumka. Archived from the original (http://arti
cles.gmdh.net/theory/bookNoiseIm.pdf) (PDF) on 2017-12-31. Retrieved 2019-11-18.
7. Ivakhnenko, O.G.; Lapa, V.G. (1967). Cybernetics and Forecasting Techniques (https://archiv
e.org/details/cyberneticsforec0000ivak) (Modern Analytic and Computational Methods in
Science and Mathematics, v.8 ed.). American Elsevier.
8. Takao, S.; Kondo, S.; Ueno, J.; Kondo, T. (2017). "Deep feedback GMDH-type neural
network and its application to medical image analysis of MRI brain images". Artificial Life
and Robotics. 23 (2): 161–172. doi:10.1007/s10015-017-0410-1 (https://doi.org/10.1007%2F
s10015-017-0410-1). S2CID 44190434 (https://api.semanticscholar.org/CorpusID:4419043
4).
9. Ivahnenko, O.G. (1982). Inductive Method of Models Self-organisation for Complex Systems
(https://web.archive.org/web/20171231104130/http://articles.gmdh.net/theory/bookInductMo
del.pdf) (PDF). Kyiv: Naukova Dumka. Archived from the original (http://articles.gmdh.net/the
ory/bookInductModel.pdf) (PDF) on 2017-12-31. Retrieved 2019-11-18.
10. Ivakhnenko, O.G.; Ivakhnenko, G.A. (1995). "The Review of Problems Solvable by
Algorithms of the Group Method of Data Handling (GMDH)" (http://www.gmdh.net/articles/rev
iew/algorith.pdf) (PDF). Pattern Recognition and Image Analysis. 5 (4): 527–535.
CiteSeerX 10.1.1.19.2971 (https://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.19.29
71).
11. Sohani, Ali; Sayyaadi, Hoseyn; Hoseinpoori, Sina (2016-09-01). "Modeling and multi-
objective optimization of an M-cycle cross-flow indirect evaporative cooler using the GMDH
type neural network". International Journal of Refrigeration. 69: 186–204.
doi:10.1016/j.ijrefrig.2016.05.011 (https://doi.org/10.1016%2Fj.ijrefrig.2016.05.011).
12. Li, Rita Yi Man; Fong, Simon; Chong, Kyle Weng Sang (2017). "Forecasting the REITs and
stock indices: Group Method of Data Handling Neural Network approach". Pacific Rim
Property Research Journal. 23 (2): 123–160. doi:10.1080/14445921.2016.1225149 (https://d
oi.org/10.1080%2F14445921.2016.1225149). S2CID 157150897 (https://api.semanticschol
ar.org/CorpusID:157150897).

External links
Library of GMDH books and articles (http://gmdh.net/articles/index.html)
Group Method of Data Handling (http://gmdh.net)

Further reading
A.G. Ivakhnenko. Heuristic Self-Organization in Problems of Engineering Cybernetics (http://
www.gmdh.net/articles/history/heuristic.pdf), Automatica, vol.6, 1970 — p. 207-219.
S.J. Farlow. Self-Organizing Methods in Modelling: GMDH Type Algorithms. New-York,
Bazel: Marcel Decker Inc., 1984, 350 p.
H.R. Madala, A.G. Ivakhnenko. Inductive Learning Algorithms for Complex Systems
Modeling (http://gmdh.net/articles/theory/GMDHbook.pdf). CRC Press, Boca Raton, 1994.

Retrieved from "https://en.wikipedia.org/w/index.php?title=Group_method_of_data_handling&oldid=1123772814"