Name
Lutsenko Yevgeniy Veniaminovich
Scholastic degree
•
Academic rank
professor
Honorary rank
—
Organization, job position
Kuban State Agrarian University
Web site url
Articles count: 271
The article discusses the application of automated system-cognitive
analysis (ASC-analysis), its mathematical
model which is system theory of information and
its software tool, which is intellectual system called
"Eidos" for solving problems related to identification
of types and models of aircraft by their silhouettes on
the ground, to be more precise, their external contours:
1) digitization of scanned images of aircraft and creation
of their mathematical models; 2) formation of
mathematical models of specific aircraft with the use
of the information theory; 3) modeling of the generalized
images of various aircraft types and models and
their graphic visualization; 4) comparing an image of a
particular plane with generalized images of various
aircraft types and models, and quantifying the degree
of similarities and differences between them, i.e., the
identification of the type and model of airplane by its
silhouette (contour) on the ground; 5) quantification of
the similarities and differences of the generalized images
of the planes with each other, i.e., clusterconstructive
analysis of generalized images of various
aircraft types and models. The article gives a new approach
to digitizing images of aircraft, based on the
use of the polar coordinate system, the center of gravity
of the image and its external contour. Before digitizing
images, we may use their transformation, standardizing
the position of the images, their sizes (resolution,
distance) and the angle of rotation (angle) in three dimensions.
Therefore, the results of digitization and
ASC-analysis of the images can be invariant (independent)
relative to their position, dimensions and
turns. The shape of the contour of a particular aircraft
is considered as a noise information on the type and
model of aircraft, including information about the true
shape of the aircraft type and its model (clean signal)
and noise, which distort the real shape, due to noise
influences, both of the means of countering detection
and identification, and environment. Software tool of
ASC-analysis, i.e. Eidos intellectual system, provides
identification of the type and the model of airplane by
its silhouette, as it was shown in a simplified numerical
example
The article discusses the application of automated system-cognitive
analysis (ASC-analysis), its mathematical
model is a system of information theory and implements,
its software tools – intellectual system called
"Eidos" for solving one of the important tasks of ampelography:
to quantify the similarities and differences
of different clones of grapes using contours of the
leaves. To solve this task we perform the following
steps: 1) digitization of scanned images of the leaves
and creation their mathematical models; 2) formation
mathematical models of specific leaves with the application
of information theory; 3) modeling the generalized
images of leaves of different clones on the basis
of specific leaves (multiparameter typing); 4) verification
of the model by identifying specific leaf images
with generic clones, i.e., classes (system identification);
5) quantification of the similarities and differences
of the clones, i.e. cluster-constructive analysis of
generalized images of leaves of various clones. The
specific shape of the contour of the leaf is regarded as
noise information on the clone to which it relates, including
information about the true shape of a leaf of
this clone (clean signal) and noise, which distort the
real shape, due to the random influence of the environment.
Software tools of ASA-analysis which is
intellectual "Eidos" system provides the noise suppression
and the detection of a signal about the true shape
of a leaf of each clone on the basis of a number of
noisy concrete examples of the leaves of this clone.
This creates a single image of the shape of the leaf of
each clone, independent of their specific implementations,
i.e. "Eidos" of these images (in the sense of Plato)
- the prototype or archetype (in the Jungian sense)
of the images
Classic quantitative measure of the reliability of the models: F-measure by van Rijsbergen is based on counting the total number of correctly and incorrectly classified and not classified objects in the training sample. In multiclass classification systems, the facility can simultaneously apply to multiple classes. Accordingly, when the synthesis of the model description is used for formation of generalized images of many of the classes it belongs to. When using the model for classification, it is determined by the degree of similarity or divergence of the object with all classes, and a true-positive decision may be the membership of the object to several classes. The result of this classification may be that the object is not just rightly or wrongly relates or does not relate to different classes, both in the classical F-measure, but rightly or wrongly relates or does not relate to them in varying degrees. However, the classic F-measure does not count the fact that the object may in fact simultaneously belongs to multiple classes (multicrossover) and the fact that the classification result can be obtained with a different degree of similarity-differences of object classes (blurring). In the numerical example, the author states that with true-positive and true-negative decisions, the module similarities-differences of the object classes are much higher than for false-positive and false-negative decisions. It would therefore be rational to the extent that the reliability of the model to take into account not just the fact of true or false positive or negative decisions, but also to take into account the degree of confidence of the classifier in these decisions. In the intellectual system called "Eidos", which is a software toolkit for the automated system-cognitive analysis (ASC-analysis), we use initially proposed by its developers measure of the reliability of the models, which is essentially a fuzzy multiclass generalization of the classical F-measure (it is proposed to call it the L-measure). In this article, L-measure is mathematically described and its application is demonstrated on a simple numerical example
The article proposes to use the automated systemcognitive
analysis (ASC-analysis) and its software tool
which is "Eidos" system to solving multiparameter
typing, system identification and cartographic visualization
of spatially-distributed natural, environmental
and socio-economic systems. Imagine, that we have an
original point cloud with coordinates (X,Y,Z), each
with known values of gradation descriptive scales of
nominal, ordinal, or numeric type S(s1,s2,...,sn). Then
the "Eidos" system provides: 1) building a model that
contains generalized knowledge about the strength and
the direction of the influence of descriptive gradations
of scales at Z=M(S); 2) estimation of the values of Z
for points (X,Y) described in the same descriptive
scales S(s1,s2,...,sn), but not a part of the original point
cloud; 3) a cartographic visualization of the spatial
distribution of values of the function Z=M(S) for
points outside the initial cloud, using Delaunay triangulation.
Basically, this means that the "Eidos" system
ensures recovery of the unknown function values on
the grounds of the argument and implements it in a
generic setting, independent of subject area. We propose
a new scientific concept called "Geo-cognition
system", which is defined as a software system that
provides conversion of source data into information,
and knowledge in visualization and mapping of this
knowledge, resulting in the cognitive map becomes
graphics. This feature can be used to quantify the degree
of suitability of the watersheds for cultivation of
certain crops, the evaluation of the ecological situation
on particular territories on the structure and intensity
of anthropogenic load, visualization of results of forecasting
of earthquakes and other unwanted risks or
emergencies, as well as for solving many other similar
mathematical essence of tasks in a variety of subject
areas. We have also shown a simple numerical example
To increase the validity of conclusions about the impact of the environment on the quality of life we need to move from generalities to the application of quantitative modeling techniques. This requires the joint processing environmental databases and databases depicting various aspects of quality of life. These databases are needed to be handled not just together, but in a comparable form approach, including technology and methodology, and to be implemented in one software system. For the first time in the environmental studies, it has been planned to be done with the application of the ASK-analysis and the system called "Eidos". Previously, the authors have set the goals and the objectives of the application of the ASK-analysis to study the effect of environmental factors on the quality of life of the population of the region. The article reveals the urgency of this study; the requirements for the method of conducting the study, the choice of a research method; as well as the contents of the objectives of the study. The proposed work is at the edge of mathematical ecology and mathematical modeling of quality of life (which refers to mathematical and instrumental methods of Economics), resulting from expected synergies, consists in obtaining of new knowledge in these fields that is relevant to both ecology and economy. This knowledge will make it more meaningful and justified for the application of environmental criteria and concepts in the economy. This work contains a description of the basic data sources for the study of the impact of environmental factors on various aspects of quality of life of the region's population, the source data for this study, the characteristics of the original data, substantiation of requirements to the method of research, choosing research methods appropriate to requirements; the development of steps to achieve the objectives of the study
HOW TO SOLVE THE TASK OF CLASSIFICATION OF TYPES OF RIFLE AMMUNITION USING THE METHOD OF ASCANALYSIS
In criminology, there are actual problems of determining
the type (machine gun, rifle, large caliber, pistol)
and a particular model of small rifle for its ammunition,
in particular, discovered in the use of weapons.
The article proposes a solution to this problem with the
use of a new innovative method of artificial intelligence:
automated system-cognitive analysis (ASCanalysis)
and its programmatic toolkit – a universal
cognitive analytical system called "Eidos". In the system
of "Eidos", we have implemented a software interface
that provides input to the system images, and the
identification of their external contours on the basis of
luminance and color contrast. Typing by multiparameter
contour images of specific ammunition, we create
and verify the system-cognitive model, with the use of
which (if the model is sufficiently reliable), we can
solve problems of system identification, classification,
study of the simulated object by studying its model
and others. For these tasks we perform the following
steps: 1) enter the images of ammunitions into the system
of "Eidos" and create mathematical models of
their contours; 2) synthesis and verification of models
of the generalized images of ammunition for types of
weapons based on the contour images of specific munitions
(multivariate typology); 3) quantification of the
similarities-differences of the specific ammunition
with generalized images of ammunition of various
types and models of small rifle (system identification);
4) quantification of the similarities-differences of the
types of munitions, i.e. cluster-constructive analysis
Classical combinatorial formula to calculate the number
of combinations from n on m: C(n,m)=n!/(m!(nm)!)
involves the intermediate calculation of factorials,
which is often impossible when n>170, due to limitations
in the capacity of numbers that are used in programming
languages and created through these systems.
However, in some cases it is necessary to calculate
the number of combinations for n and m much
larger than this limit, such as when a value greater than
10000. In such cases, there is a definite problem,
which manifests itself, for example in the fact that
many on-line services meant to calculate the number
of combinations with these parameters do not work
properly. In this article, we present its solution in the
form of an algorithm and software implementation.
The essence of the approach is to first decompose the
factorials into prime factors and reduce them, and then
to produce multiplication. This approach differs from
those cited in the Internet
There are three main growth points of modern information technologies: global network and mobile communication, advanced human-machine interfaces, intelligent technologies. As it is known, the system (synergistic) effect is usually observed in multidisciplinary and interdisciplinary researches. This means that an interesting direction of research and development is located at the overlap of these three promising areas, namely: advanced interfaces in the global mobile networks, advanced intelligent interfaces and the application of artificial intelligence technologies in the Internet and mobile communications. In addition, a particularly high relevance goes to the development and application prospective of intelligent interfaces to the Internet and mobile communications. The Internet intellectualities gradually, it turns from non-local storage of large data (big data) in information space that contains meaningful big data, i.e. "great information" (great info), and then in the space of knowledge or "cognitive space" in which most information is actively used to achieve goals (management) and turns into the "great knowledge" (great knowledge). There are more sites devoted to artificial intelligence, free databases for machine learning (UCI, Kaggle, and others) and even on-line intelligent applications, and interfaces used in the Internet are improving. Recently, there was an acquisition of company Oculus, which is the world's leading developer and manufacturer of ammunition of virtual reality by the developer of one of the first global social networking Facebook - Mark Zuckerberg. However, students and scientists still do not notice that open, scalable, interactive, intelligent on-line environment for learning and researches already exists and operates, based on automated system-cognitive analysis (ASC-analysis) and its programmatic Toolkit – intellectual "Eidos" and the author's website. This article is an original presentation and it is designed to familiarize potential users with the capabilities of this environment
In this article, in accordance with the methodology of
the Automated system-cognitive analysis (ASCanalysis),
we examine the implementation of the 3rd
ASC-analysis: synthesis and verification of forecasting
models of development of diversified agro-industrial
corporations. In this step, we have synthesis and verification
of 3 statistical and 7 system-cognitive models:
ABS – matrix of the absolute frequencies, PRC1 and
PRC2 – matrix of the conditional and unconditional
distributions, INF1 and INF2 private criterion: the
amount of knowledge based on A. Kharkevich, INF3 –
private criterion: the Chi-square test: difference between
the actual and the theoretically expected absolute
frequencies INF4 and INF5 – private criterion:
ROI - Return On Investment, INF6 and INF7 – private
criterion: the difference between conditional and unconditional
probability (coefficient of relationship).
The reliability of the created models was assessed in
accordance with the proposed metric is similar to the
known F-test, but does not involve the performance of
normal distribution, linearity of the object modeling,
the independence and additivity acting factors. The
accuracy of the obtained models was high enough to resolve the subsequent problems of identification,
forecasting and decision making, as well as studies of
the modeled object by studying its model, scheduled
for consideration in future articles
The article considers the application of Eidos intellectual technologies for implementation of developed veterinary and medical diagnostics statistical tests without programming in the convenient form for the individual and mass testing, the analysis of the results and development of the individual and group recommendations. It is possible to merge several tests in one supertest