PRODUCTS DEVELOPED BY ADUIS
Confor
TOOLS FOR KNOWLEDGE
DISCOVERY, CLASSIFICATION,
DIAGNOSTICS AND PREDICTION.
Analogy
TOOLS FOR SOLVING
PROBLEMS ON THE BASIS OF ANALOGY
Manager
TOOLS FOR DECISION
SUPPORT SYSTEMS DESIGN
Discret
TOOL FOR DISCRETIZATION
OF NUMERICAL DATA
Gobsec
THE SYSTEM FOR
INVESTMENT SCHEDULING
TOOLS FOR KNOWLEDGE DISCOVERY,
CLASSIFICATION,
DIAGNOSTICS AND PREDICTION.
The CONFOR
system supports the most effective methods of intelligent data analysis.
The main
functions of CONFOR are:
- revealing the regularities that
characterize classes of objects which are represented as sets of attribute
values
- using the revealed regularities for
classification, diagnostics and prediction.
Original methods
of knowledge discovery based on the pyramidal networks are taken as a principle
in the CONFOR system. A successful long-term application of the
methods in different fields of research and development has confirmed their decisive
advantages as compared to other known methods.
Chemists have
done over a thousand of high-precise prognoses for new chemical compounds and
materials. CONFOR is used for analysis of information
technologies market. Application field for CONFOR is also
medicine, economy, ecology, geology, technical diagnostics, sociology, etc.
Pyramidal
networks considerably decrease volumes of search operations that makes it
possible to avoid the effect of "information explosion" when solving
analytical problems on the basis of large-scale data.
The system forms
logical models of object classes which allow to take into account influence on
diagnosed or prognosed parameter of separate attributes and their various
combinations as well. Besides, it is taken into account influence of
"exclusive" attributes, that is such attributes which are
incompatible with diagnosed or prognosed value. These properties of models of
object classes favour increasing diagnosis and prognosis accuracy.
CONFOR is able to discover a
regularity however complex it be.
THE INITIAL
DATA for
the CONFOR system are text files (format csv) or dbf tables.
SYSTEM
REQUIREMENTS. The CONFOR system requires operational system WINDOWS,
processor type PENTIUM, 20 Mb free space on a hard
drive.
The
slideshow on the Confor system
you can take here.
Load file CnfEng.exe (821Kb). It is self-extracting archive. It contains the
presentation CnfEng.ppt.
here.
REFERENCES
- V.P.Gladun, Z.L.Rabinovich.
"Formation of the World Model in Artificial Intelligence
Systems." In: Machine Intelligence, 9, Ellis Herwood Ltd.,
Chichester, 1980, pp. 299-309.
- V.P.Gladun. Planning of Solutions. Kiev: Naukova Dumka,
1987. - 186 p. (in Russian).
- V.P.Gladun. Processes of New Knowledge Formation. Sofia: CD
Pedagog, 1994. - 192 p. (in Russian).
- V.P.Gladun. Partnership with a Computer. - Kiev: Port-Royal,
2000. - 128 p. (in Russian).
- V.P.Gladun, N.D.Vashchenko. Local
Statistical Methods of Knowledge Formation// Kibernetika i Sistemni
analiz. 1995, N2, pp. 62-74. (in Russian).
- Kiselyova N.N., Gladun V.P., Vashchenko
N.D., LeClair S.R., Jackson А.G. Prediction of Inorganic Compounds Perspective for Search of New
Electrooptical Materials// Perspektivnie Materiali, 1998, N 3.
pp.28 -32. (in Russian).
- Gladun V., Vashchenko N. Local Statistical
Methods of Knowledge Formation// Cybernetics and System Analysis.-1995.-v.31.-N2.-P.207-217.
- N.N.Kiselyova, V.P.Gladun, N.D.Vashchenko.
"Computational Materials Design Using Artificial Intelligence
Methods". Journal of Alloys and Compounds, 279 (1998),
pp.8-13.
TOOLS FOR SOLVING PROBLEMS ON THE
BASIS OF ANALOGY
The ANALOGY
system is aimed at:
- desing of complex objects with predefined
properties;
- desing of complex objects that are
analogous to specified ones;
- determination of unknown properties of
complex objects;
- search for objects that are analogous to
ones specified in data bases.
The system can
be adapted to different application domains by means of filling the appropriate
database. The database of the system includes descriptions of:
- known properties of complex objects;
- composition of complex objects;
- properties of elements that are
constituents of complex objects;
- rules for constructing complex objects.
The ANALOGY
system provides:
- filling the data base in a dialogue mode
using the limited natural language;
- filling the data base from files;
- adaptation of the man-machine dialogue to
the application domain;
- solution of problems on the basis of
analogy.
The ANALOGY
system makes it possible the following knowledge analysis:
- revealing and estimating the analogy
degree of the given objects;
- finding both common and specific features
in object descriptions;
- control of completeness and contradiction.
Typical
application fields for the ANALOGY system are chemistry, material
science, genetics, economy.
SYSTEM
REQUIREMENTS. Operational system WINDOWS, processor type PENTIUM, Borland Database
Engine 5.01 or higher; 7Mb additional free space on a hard drive.
REFERENCES
- V.P.Gladun. Partnership with a Computer. - Kiev: "Port-Royal",
2000. - 128 p. (in Russian).
- V.P.Gladun, V.Ju.Velichko, N.N.Kiselyova,
N.N.Moskalkova. Inference of Hypothesis on Composition and Properties of
Objects on the Basis of Analogy //Iskusstvennij Intellekt. - 2000.
- v 1. - pp. 44-52.(in Russian).
TOOLS FOR DECISION SUPPORT SYSTEMS
DESIGN
MANAGER tools are to design
application systems intended for situation analysis and multi-criteria decision
making when controlling complex industrial, administrative or technical
objects.
MAIN
FUNCTIONS:
- storage the information characterizing
situation in the controlled sphere at every decision-making step;
- analysis of a current situation and warning
about deviations of its characteristics from the norm;
- formation of recommendations concerning
elimination of deviations;
- evaluation of recommended alternative
decisions by means of modeling their multi-step consequences.
The tools can be
used for support of decision making process in abnormal situations when
controlling complex technical objects; with the aim of technical diagnostics;
for management of firms and administrative regions; for manager training, etc.
MANAGER tools makes it possible
to choose decisions under condition when mapping of alternatives of a decision
into their criteria estimations is not given directly.
The domain of MANAGER
application must be represented with sets of parameters, relations among them
and permittable controlling actions. On the basis of this information MANAGER
form a model of application domain which can be modified if the controlled
object is changed. When choosing decisions MANAGER tools makes
it possible to take into account their multi-step consequences, as distinct
from another tools of decision making support.
REFERENCES
- V.P.Gladun. Planning of Solutions. Kiev: Naukova Dumka,
1987. - 186 p. (in Russian).
- Gladun V., Vaschenko N. Control on the
Basis of Network Models. Artificial Intelligence in Real-Time Control.
7th IFAC Symposium, Arizona,USA, 1998, pp.88-93.
- Gladun V.P., Vaschenko N.D. et el. Balance
Methods of Action Planning in Control Systems. Upravljajushchie
Sistemi i Mashini. Kiev, N1/2, 1995, pp.93-96 (in Russian).
TOOL FOR DISCRETIZATION OF NUMERICAL
DATA
The system is
intended for comparative analysis of numerical characteristics of objects which
belong to different classes, as well as for transformation of numerical
attributes to nominal ones on the stage of data preprocessing when solving
discrete analytical problems (knowledge discovery, classification, diagnostics,
prediction).
The system can
be used in different application domains, for example, for:
- comparative estimation of technical and
market characteristics of different products;
- discretization and comparison of
physical-chemical characteristics of various materials;
- comparison of numerical parameters which
characterize enterprises or regions in different time periods and so on.
The system
divides scales of numerical parameters into intervals comparing their value
distributions which characterize different classes of objects. As a result
intervals appear that to the most degree characterize the investigated classes
of objects. Application of the system for data preprocessing increases the
accuracy of classification, diagnostics or prediction.
The system can
interface with the CONFOR system intended for
intelligent data analysis. The peculiarity of the system, as compared to
another systems of the same purpose, are:
- wholly automated discretization process;
- possibility to select intervals taking
into account distributions of objects of different classes that enables to
form more effective classification rules when solving discrete analytical
problems.
THE INPUT AND
OUTPUT DATA for the system are text files.
SYSTEM
REQUIREMENTS. Operational system WINDOWS, processor type PENTIUM, 15 Mb additional free
space on a hard drive.
THE SYSTEM FOR INVESTMENT SCHEDULING
PURPOSE.
Success of your
business sometimes depends to a great degree on its versatility. Investments in
different fields of economy can help you to keep and increase your profit even
under conditions of market instability. In this case it is necessary to
calculate in a proper way the sequence of investments for different kinds of
business to provide the maximum profit.
The GOBSEC
system forms the sequence of projects to be fulfilled that provides the maximum
profit taking into account necessary investments and repayments.
WHAT TASKS
DOES GOBSEC SOLVE?
Let us consider
a situation of scheduling of financial operations. Let you have a sum for
investment (investment fund) and are going to fulfil some operations (to buy or
to sell goods, to credit somebody, to invest a project, to do some service, to
publish a book and so on).
You should know
for every operation:
- the moments of the necessary investments
relative to the beginning of the operation;
- the money sum to be invested at every
moment;
- the moments of repayments relative to the
beginning of the operation;
- the repayment sums at every moment;
- the expected profit during a definite time
after the operation completion.
Now let us try
to solve a task: when should we begin every operation taking into account
necessary investment and repayment sums, if we want to have the maximum profit
for all operations in total. The task is rather complicated, if the initial
investment fund is unsufficient to provide simultaneous fulfilment of all
operations. Use your practice to consider the above task for 5-6 operations,
and you will see that even for such low number of operations it is rather
difficult to find the most profitable investment strategy.
These are the
very tasks that GOBSEC can solve.
GOBSEC will simplify for you
complicated combinative calculations and prompt you what to do. You can correct
the invesment plan that was produced by GOBSEC. You can
delete an operation from the GOBSEC's plan or move its
beginning to a later date. GOBSEC will offer you another plan
taking into account your corrections.
Making various
corrections, it is possible to construct different plan variants. GOBSEC
will save them and show at the end of your work together with the number of
projects that can be invested and some additional information.
USERS OF THE
SYSTEM:
- firms with versatile business;
- innovation funds;
- banks and insurance companies;
- educational institutions.
ADUIS | Publications | Conferences | Write to us
|
|
|
Copyright (c) 1998-2018 ADUIS www.aduis.com.ua |