Alexey Grigorevich Ivakhnenko
Збережено в:
| Дата: | 2013 |
|---|---|
| Автор: | |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Міжнародний науково-навчальний центр інформаційних технологій і систем НАН та МОН України
2013
|
| Назва видання: | Індуктивне моделювання складних систем |
| Теми: | |
| Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/83659 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | Alexey Grigorevich Ivakhnenko / M. Snorek // Індуктивне моделювання складних систем: Зб. наук. пр. — К.: МННЦ ІТС НАН та МОН України, 2013. — Вип. 5. — С. 87-89. — Бібліогр.: 2 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-83659 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-836592015-06-22T03:02:11Z Alexey Grigorevich Ivakhnenko Snorek, M. Спогади про Олексія Григоровича Івахненка 2013 Article Alexey Grigorevich Ivakhnenko / M. Snorek // Індуктивне моделювання складних систем: Зб. наук. пр. — К.: МННЦ ІТС НАН та МОН України, 2013. — Вип. 5. — С. 87-89. — Бібліогр.: 2 назв. — англ. XXXX-0044 http://dspace.nbuv.gov.ua/handle/123456789/83659 en Індуктивне моделювання складних систем Міжнародний науково-навчальний центр інформаційних технологій і систем НАН та МОН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| topic |
Спогади про Олексія Григоровича Івахненка Спогади про Олексія Григоровича Івахненка |
| spellingShingle |
Спогади про Олексія Григоровича Івахненка Спогади про Олексія Григоровича Івахненка Snorek, M. Alexey Grigorevich Ivakhnenko Індуктивне моделювання складних систем |
| format |
Article |
| author |
Snorek, M. |
| author_facet |
Snorek, M. |
| author_sort |
Snorek, M. |
| title |
Alexey Grigorevich Ivakhnenko |
| title_short |
Alexey Grigorevich Ivakhnenko |
| title_full |
Alexey Grigorevich Ivakhnenko |
| title_fullStr |
Alexey Grigorevich Ivakhnenko |
| title_full_unstemmed |
Alexey Grigorevich Ivakhnenko |
| title_sort |
alexey grigorevich ivakhnenko |
| publisher |
Міжнародний науково-навчальний центр інформаційних технологій і систем НАН та МОН України |
| publishDate |
2013 |
| topic_facet |
Спогади про Олексія Григоровича Івахненка |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/83659 |
| citation_txt |
Alexey Grigorevich Ivakhnenko / M. Snorek // Індуктивне моделювання складних систем: Зб. наук. пр. — К.: МННЦ ІТС НАН та МОН України, 2013. — Вип. 5. — С. 87-89. — Бібліогр.: 2 назв. — англ. |
| series |
Індуктивне моделювання складних систем |
| work_keys_str_mv |
AT snorekm alexeygrigorevichivakhnenko |
| first_indexed |
2025-07-06T10:28:32Z |
| last_indexed |
2025-07-06T10:28:32Z |
| _version_ |
1836893038784282624 |
| fulltext |
Спогади про О.Г. Івахненка
ALEXEY GRIGOREVICH IVAKHNENKO
Miroslav Snorek
Czech Technical University in Prague,
Prague, Czech Republik
With the name Alexey Grigorevich Ivakhnenko I met firstly in eighties
(last century). We were collecting at that time materials for the university course
Artificial Neural Networks. The main impulse for my deeper studies of his
Group Method of Data Handling (GMDH)was a notice that they are “self-
organizing”.Later on I found that this property means something else, than I
expected at that time. But later on we have started our own research in Data
Mining based on original Ivakhnenko’s ideas.
Itis known that GMDH nets are based on induction. Not going into the
details (see the original GMDH literature, e.g. [1]), they search by different
heuristics (COMBI, MIA) for most promising subspaces in the space of possible
solutions. Through self-organizing they improve the reached solution quality. In
the case that reached solutiondoes not meet the expectations, the net topology
changes (a new layer of processing element is added). This process follows
subsequently whether even the new solution does not generate the expected
solution quality. Processing elements in a particular layer of the GMDH
architecture are equal in each layers, it means they are of the same
approximation polynomial order (MIA).In the other case (COMBI) new units do
apply higher order approximation polynomials to the original inputs.
Індуктивне моделювання складних систем, випуск 5, 2013 87
Miroslav Snorek
In Dresden in 1993.On the picture (from left):
G. A. and A. G. Ivakhnenko, Czech scientist M. Jirina
Our first contacts were only passive, but this situation has to be changed
soon. At the beginning we were only professionally motivated by GMDH
literature sources. In 1993 we, it means CTU University in Prague people
together with Czech Academy of Sciences colleagues, have started to prepare a
common scientific project with people from Ukrainian Academy of Science.
The key person was German professor Johan-Adolf Mueller affiliated at that
time in Dresden at FTW (FRG). We found soon that Johan-Adolf was
supervised by A.G. Ivakhnenko earlier in his PhD study time. In Dresden that
was my first time personal contact with Alexey Grigorevich. I remember well,
how unforgettable my discussions with himwere. Those several days of common
grant proposal preparations were really nice. A great pity not only for me was
the fact that the proposal was not accepted.
The second of my professional contacts was indirect, but also intensive.
One of my doctoral students, namely PavelKordik, improved his study time by
international experience stay in the New Zeeland. His activities there were not
only enlarging his intellectual level. He started there as mandarin fruit harvester.
This activity as I understood from his later given stories is mostly manual, hard
and monotonous. One has enough time for thinking and therefore no doubt that
Pavelnoticed a local guy which activity was more interesting. He was collecting
input data for water supply modeling of mandarin tree growth and fruit
production. The goal of this activity was possible improvement of the harvest
quantity with water consumption optimization. Based on the ideas exchanged a
common research project started. In fact here started the first ideas which led to
the formulation of Kordik‘s GAME neural nets [2]. Dissertation was
successfully defended in 2007. Pavel visited several times the Ukrainetop
GMDH investigating home place to let know results of his performed research.
Here he often met Alexey Grigorijevich.
A.G. Ivakhnenko, father of the discipline,was actively visiting scientific
events concerning inductive modeling not only in Ukraine. Even in his high age
he was brilliant in discussions with his young colleagues. I personally admire his
overview and knowledge about who is interested in what.
My next, but unfortunately last personal contact with Alexey
Grigorievich, was during the European Simulation Multiconference ESM in
Індуктивне моделювання складних систем, випуск 5, 2013 88
Спогади про О.Г. Івахненка
2001 in Prague, Czech Republic. We invited him as Invited Speaker for the
Artificial Neural Network Conference. Despitehe was nearly 90, we were
pleasantly surprised obtaining his enrolment and later of his arrival accompanied
by his son Gregory.
I do remember as to be today that the official Conference Dinner took
place in Cabaret of the famous Beerpub „U Fleku“. The atmosphere was there
really friendly. All over 200 participants were close to the Midnight cheerful,
but alsotired. Therefore a surprise for me as the Conference Chair was that
someone invited me to prolong the event on a private party. Shocking was that
the invitation was send to me by Alexey Grigorievich and it was presented to me
by his next formerly doctoral lady student. So we together with my wife
achieved late in the night in the private flat above the Vltava river, where lived
Vaclav Havel family. The hosting lady was the wife of younger of them, Ivan.
He belongs by the way to the top of IT experts not only among Czech Republic
residents. We later understood that all approximately 20 late-night party
members are former Alexey Grigoriewich PhD students now living in different
parts of the world.
Imprints of Ivakhnenko’sintellectual capital are clearly evident in many
today used machine learning methods. He really strongly influenced not only his
students, bud he succeeded to enforce his
Literature
[1] http://www.gmdh.net/articles/index.html
[2] http://cig.felk.cvut.cz/research/publications/kordik-dizertace-1.pdf
Індуктивне моделювання складних систем, випуск 5, 2013 89
|